Pith. sign in

REVIEW 3 major objections 6 minor 300 references

Sovereign Large Language Models: Advantages, Strategy and Regulations

T0 review · 3 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

Pith's one-line read A country can build a sovereign large language model through coordinated investment, infrastructure, and regulation, according to a survey of national LLM projects across 18 macroregions.

desk verdict Useful policy mapping of sovereign LLM programs, but the headline efficiency claim is anecdotal and overstates what the evidence shows. read the letter →

arxiv 2503.04745 v1 pith:VTT3CRLR submitted 2025-02-05 cs.CY cs.CL

classification cs.CYcs.CL
keywords sovereignAIlargelanguagemodelsnationalstrategyregulationgovernmentinvestmentEUprogramsadministrativeefficiencydatasovereignty
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This report argues that national governments can develop sovereign large language models by combining state strategy, public investment, and tailored regulation. It surveys LLM projects and policies in more than 18 macroregions, covering roughly 85 percent of the world's population, and finds recurring patterns: countries fund models through state budgets, corporate ventures, grants, or hybrid arrangements; they regulate usage through access restrictions, licensing, auditing, and risk-based rules; and they pursue strategic autonomy, data security, and cultural or linguistic preservation. The stated payoff is administrative efficiency: government adoption of AI reportedly grew more than 1.5 times between 2021 and 2024, with uses in legal document handling, citizen services, healthcare, and defense. The report's practical target is Ukraine, which the authors argue can follow the same path using EU programs such as Horizon Europe, Digital Europe, and EuroHPC.

What carries the argument

The central mechanism is the comparative case survey: the authors systematically ask, for each country, who trains models, how training and use are regulated, how training data is regulated, how training is funded, which institutions are involved, and which international programs they joined. The answers are grouped into categories that form the report's analytical skeleton: strategic goals (autonomy, ethics, infrastructure, data access, talent), funding sources (state, business, grants, hybrid), and regulatory modes (access restrictions, licensing, monitoring, responsibility allocation, risk-based rules). This categorization, not a formal model or dataset, carries the argument: it converts many individual examples into transferable patterns.

What would settle it

A re-analysis that includes low-income countries and independently verifies the cited investment figures would settle the claim; if LLM adoption does not correlate with the claimed administrative-efficiency gains across countries, or if key numbers such as Brazil's $4.2 billion commitment or the EU's €1 billion annual AI investment turn out to be misstated, the report's broad conclusions would not follow.

Watch

Extended reading notes

Core claim

The paper's central claim is that sovereign LLMs are both desirable and achievable for countries outside the US-China axis, provided the state coordinates investment, infrastructure, and regulation. On the evidence assembled, national models outperform global models on localized benchmarks and offer data protection, national-security alignment, and cultural preservation. The report catalogs concrete projects: BgGPT in Bulgaria, GPT-NL in the Netherlands, GPT-SW3 in Sweden, ALLaM in Saudi Arabia, SEA-LION in Singapore, Sabiá in Brazil, and Fugaku-LLM in Japan, among others. It concludes that hybrid funding combining state, corporate, and grant money is common, and that regulatory frameworks are converging on risk-based oversight, transparency, and licensing for sensitive sectors. The argument is inductive: repeated country cases are meant to show that the category 'sovereign LLM' is a working policy option, not a theoretical one.

Load-bearing premise

The report assumes that the countries and projects it selected for analysis, drawn mainly from high- and middle-income nations and documented through press releases, news articles, and vendor blogs, are representative enough to support general conclusions about global sovereign-LLM trends.

Editorial extensions

If this is right

  • Countries with limited AI budgets can pursue sovereign LLMs through hybrid funding, as seen in Thailand's OpenThaiGPT and Greece's Faros project.
  • EU programs, including Horizon Europe, Digital Europe, and EuroHPC, are accessible to Ukraine and can supply grants, supercomputing time, and data infrastructure for training national models.
  • Regulatory convergence is moving toward risk-based oversight: banning unacceptable uses, licensing high-risk applications, and auditing government AI, as illustrated by the EU AI Act, Thailand's Royal Decree, Brazil's Bill 2,338/2023, and South Korea's AI Basic Act.
  • National models can outperform general-purpose global models on local benchmarks and better protect citizens' data, making sovereignty a concrete performance and security choice rather than only a symbolic one.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's sample is weighted toward high- and middle-income countries and documented largely through official announcements, press releases, and vendor blogs, so its conclusions about global trends are only as strong as that convenience sample.
  • A rigorous causal test would compare administrative efficiency in matched governments with and without sovereign LLM deployment; the report itself offers correlations and case examples, not such a test.
  • If the pattern holds, the same playbook could apply beyond Ukraine to other mid-sized economies seeking AI independence, though the report does not quantify minimum viable budgets, compute requirements, or the time horizons involved.
  • One testable extension is to track whether countries that follow the hybrid-funding and risk-based-regulation pattern achieve faster or more sustainable LLM deployment than those relying on single-source funding or broad bans.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The manuscript is a policy report on the global development of sovereign large language models (LLMs), surveying national strategies, funding mechanisms, regulatory frameworks, and international cooperation programs, with a particular focus on what Ukraine can learn from these experiences. It contains an extensive compilation of country examples (including Albania, Brazil, Japan, the Netherlands, Saudi Arabia, Singapore, and others), investment figures, and EU programs such as CLARIN, Horizon Europe, Digital Europe, and EuroHPC. The report's central claim, stated in the Executive Summary and repeated in the conclusions, is that 'International experiences indicate that LLMs significantly enhance administrative efficiency' and that coordinated state investment, infrastructure, and regulation can enable sovereign LLMs that protect data security, support national interests, and preserve cultural and linguistic diversity.

Significance. If its central claim were well-supported, the report would be a useful mapping of sovereign-LLM initiatives and a practical guide for countries such as Ukraine considering similar investments. Its main strengths are the breadth of documented examples, the assembly of funding figures from many countries, and the careful enumeration of EU programs accessible to Ukraine. The report is also transparent in citing its sources, which number over 300. However, the report is not a systematic empirical study: it provides no measured outcome data, no comparison groups, and no robustness checks. The headline benefit claim is therefore not established by the evidence presented, and the policy recommendation built on that claim is correspondingly vulnerable.

major comments (3)
  1. [Executive Summary; §4.1] The central claim that 'International experiences indicate that LLMs significantly enhance administrative efficiency' is not supported by the evidence in the report. The examples in §4.1 (Albania's Virtual Assistant, the Netherlands' Codi, North Macedonia's ADA, Serbia's COMtext.SR, and others) are descriptive vignettes; no cited source provides a baseline, a control group, or a measured outcome such as time saved, cost reduction, error rates, or user satisfaction. The TNO finding cited for the '1.5 times' increase measures adoption counts, not efficiency gains. Because this benefit claim is the foundation for the report's recommendation that Ukraine invest in sovereign LLMs, the central inference of the report does not follow from its evidence.
  2. [§3 Methodology] The selection of countries and projects is a convenience sample of high- and middle-income countries drawn mainly from press releases, news articles, and vendor blogs. The statement that 'more than 18 macroregions were analyzed, covering about 85% of the world's population' is not accompanied by a description of how macroregions were defined, how coverage was computed, or how representativeness was assessed. Consequently, the report's general conclusions about global trends in investment, regulation, and strategic priorities, and their applicability to Ukraine, exceed what the sample can justify.
  3. [§4.3] The economic-impact statistics in §4.3 are almost entirely industry-commissioned projections, not measured outcomes, and the report does not distinguish between these categories. For example, the McKinsey figures on generative-AI contribution to the global economy, the Access Partnership/Google estimate for Asia-Pacific, the Strand Partners/AWS '€600 billion by 2030' for Europe, and the SCB EIC survey's '6% GDP by 2030' for Thailand are all forward-looking estimates that depend on assumptions the report does not evaluate. Using these projections as direct support for the claim of significant benefits conflates forecasts with established evidence.
minor comments (6)
  1. [§5.1] There is a formatting artifact 'life/emdash.cyrfrom healthcare to urban planning' that should be cleaned up, and the phrase 'success criterias' should be 'success criteria'.
  2. [§5.2.2] In the description of the GPT-NL project, 'this organization would playing a key role' is ungrammatical; it should read 'would play a key role.'
  3. [§5.2] In the paragraph on tokenizers, 'Thissegmentation process' is missing a space and should read 'This segmentation process.'
  4. [§4.1 introductory paragraph] The section promises to 'shed light on their limitations' of the LLM applications discussed, but the subsequent examples contain no actual discussion of limitations; either add such discussion or remove this promise from the introduction.
  5. [References] Several references are malformed or have duplicate numbering, for example the duplicated '[189]' in §6.2.1 and duplicated '[206]' in §6.2.2, and the reference list contains truncated URLs (e.g., [8], [86], [141]) and a typo in the AmericasNLP entry ('Thrid' for 'Third').
  6. [§6.1] The statement that the EU AI Act 'was passed in early April 2024' and 'will become effective on February 2, 2025' is imprecise: the AI Act was published in July 2024, and its obligations become applicable progressively from February 2025 onward; please check and correct the dates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the report is an evidence survey whose claims are inductive generalizations from external sources, not derivations from its own outputs.

full rationale

This paper is a policy survey and country-comparison report, not a formal derivation. It introduces no equations, no fitted parameters, no normalization choices, and no self-referential machinery. The central claim that "International experiences indicate that LLMs significantly enhance administrative efficiency" is supported by cited examples from TNO, Albania, the Netherlands, North Macedonia, and other sources; it is an inductive synthesis of external evidence, not a conclusion that was used as an input to generate itself. The methodology section discloses the sample focus on high- and middle-income countries and states that "More than 18 macroregions were analyzed, covering about 85% of the world's population," which is a potential sampling limitation but not a circular step. The skeptical concern that the report lacks baseline or outcome measurements for its efficiency claim is a correctness and evidence-quality criticism, not an instance of circular reasoning under the enumerated patterns: no quantity is defined in terms of another, no fitted value is renamed as a prediction, and no load-bearing conclusion depends on a self-citation chain. Consequently, the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No new scientific entities are introduced. The report's conclusions rest entirely on the authority of external sources, a convenience sample of country examples, and accepted industry projections.

assumptions (3)
  • domain assumption Publicly reported investment figures and regulatory descriptions are substantially accurate.
    The report relies on government press releases, news articles, and vendor blogs for numbers such as Brazil's $4.2 billion plan, the EU's €1 billion annual AI spending, and the Netherlands' €204.5 million allocation. These figures are not independently audited.
  • domain assumption The convenience sample of high- and middle-income countries is representative of global sovereign LLM trends.
    Section 3 states the authors primarily looked at high and middle-income countries and claims coverage of about 85% of the world's population, but no sampling frame or comparison to excluded countries is provided.
  • domain assumption Industry economic projections cited in Section 4.3 are valid enough to support claims about AI's economic impact.
    The report cites McKinsey's $2.6 to $4.4 trillion annual contribution estimate and Access Partnership's $3 trillion Asia-Pacific figure as evidence. These are forecasts commissioned by interested parties, not measured economic outcomes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sovereign Large Language Models: Advantages, Strategy and Regulations." pith.science (2026). https://pith.science/paper/VTT3CRLR

@misc{pith2026250304745,
  author       = {Pith},
  title        = {Pith review of: Sovereign Large Language Models: Advantages, Strategy and Regulations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VTT3CRLR}},
  note         = {Machine review of arXiv:2503.04745}
}
read the original abstract

This report analyzes key trends, challenges, risks, and opportunities associated with the development of Large Language Models (LLMs) globally. It examines national experiences in developing LLMs and assesses the feasibility of investment in this sector. Additionally, the report explores strategies for implementing, regulating, and financing AI projects at the state level.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

300 extracted references · 73 canonical work pages

  1. [1]

    U RL: https://www.ai.se/en/project/shared-digital-assista nt-public-sector

    A shared digital assistant for the public sector, 2024. U RL: https://www.ai.se/en/project/shared-digital-assista nt-public-sector

  2. [2]

    I. T. Administration. Singapore artificial intelligenc e strategy 2.0, 03 2024. URL: https://www.trade.gov/market-intelligence/singapore -artificial-intelligence-strategy-20

  3. [3]

    G. o. J. Agency for Cultural Affairs Policy. Copyright law and the use of works in education,

  4. [4]

    G. N. Agenda. Greece’s ”pharos”: A landmark proposal for europe’s ai factory initiative and for the creation of a greek large language model, 11 2024

  5. [5]

    H. AI. Foundation model. URL: https://www.hippocraticai.com/foundationmodel

  6. [6]

    S. AI. Introducing defense llama. URL: https://scale.com/blog/defense-llama

  7. [7]

    S. AI. Sarvam 1, 2024. URL: https://www.sarvam.ai/blogs/sarvam-1

  8. [8]

    URL: https://www.aboutamazon.eu/news/job-creation-and-in vestment/ai-adoption-forecast-to-unleash-600-

    Ai adoption forecast to unleash €600 billion growth in eu rope’s economy, 02 2024. URL: https://www.aboutamazon.eu/news/job-creation-and-in vestment/ai-adoption-forecast-to-unleash-600-

Show all 300 references
  1. [9]

    URL: https://www.kryeministria.al/en/newsroom/asistenti- virtual-1-0-faze-e-rendesishme-ne-rruget

    Ai-based virtual assistant 1.0, crucial stage on path to wards innovation and transformation, 2025. URL: https://www.kryeministria.al/en/newsroom/asistenti- virtual-1-0-faze-e-rendesishme-ne-rruget

  2. [10]

    Large ai grand challenge, 2023

    AI-BOOST. Large ai grand challenge, 2023. URL: https://aiboost-project.eu/large-ai-grand-challenge /

  3. [11]

    AI Israel. The israel innovation authority and israel n ational digital min- istry will fund the establishment of the association of natu ral language pro- cessing (nlp) technology companies in hebrew and arabic, 20 24. URL: https://aiisrael.org.il/press_release/the-israel-in ...

  4. [12]

    Sea-lion: South-east asia large languag e models

    AI Singapore. Sea-lion: South-east asia large languag e models. URL: https://github.com/aisingapore/sealion

  5. [13]

    Sea-lion.ai – south east asian languages in one network

    AI Singapore. Sea-lion.ai – south east asian languages in one network. URL: https://aisingapore.org/aiproducts/sea-lion/

  6. [14]

    Ai thailand annual report 2023

    AI Thailand. Ai thailand annual report 2023. Technical report, AI Thailand, 2023. URL: https://ai.in.th/wp-content/uploads/2024/06/AI-Thai land-Annual-Report-2023.pdf

  7. [15]

    About ai21 labs

    AI21 Labs. About ai21 labs. URL: https://www.ai21.com/about

  8. [16]

    Jamba: the most powerful & efficient long conte xt models

    AI21 Labs. Jamba: the most powerful & efficient long conte xt models. URL: https://www.ai21.com/jamba

  9. [17]

    French startup mistral ai raises $640m in series b t o expand its ai model development,

    AIN. French startup mistral ai raises $640m in series b t o expand its ai model development,

  10. [18]

    Al-Balushi

    R. Al-Balushi. Saudi new draft intellectual property l aw – riyadh al-balushi, 2023. URL: https://riyadh.om/2023/saudi-new-draft-intellectual -property-law/

  11. [19]

    URL: https://en.ain.ua/2024/06/13/mistral-ai-series-b-fu nding-success/

  12. [20]

    Alexandrov, V

    A. Alexandrov, V. Raychev, D. I. Dimitrov, C. Zhang, M. V echev, and K. Toutanova. Bggpt 1.0: Extending english-centric llms to other languag es, 2024. URL: https://arxiv.org/abs/2412.10893

  13. [21]

    URL: https://aleph-alpha.com

    Aleph alpha - ai for enterprises and governments, 2025. URL: https://aleph-alpha.com

  14. [22]

    Alibaba cloud’s qwen models attract ove r 90,000 enterprise adoptions within its first year, 2024

    Alibaba Cloud. Alibaba cloud’s qwen models attract ove r 90,000 enterprise adoptions within its first year, 2024. URL : https://www.alibabacloud.com/blog/alibaba-clouds-qw en-models-attract-over-90000-enterprise-adopt

  15. [23]

    Tongyi qianwen (qwen), 2023

    Alibaba Cloud. Tongyi qianwen (qwen), 2023. URL: https://www.alibabacloud.com/en/solutions/generative-ai/qwen. 34

  16. [24]

    Proceedi ngs of the third workshop on text sim- plification, accessibility and readability (tsar 2024)

    Association for Computational Linguistics. Proceedi ngs of the third workshop on text sim- plification, accessibility and readability (tsar 2024). In Proceedings of the Third Work- shop on Text Simplification, Accessibility and Readability (TSAR 2024) , 2024. URL: https://aclant...

  17. [25]

    Thrid workshop on nlp for indigenous lang uages of the americas (americasnlp),

    AmericasNLP. Thrid workshop on nlp for indigenous lang uages of the americas (americasnlp),

  18. [26]

    Atanasova

    T. Atanasova. Bulgaria becomes the first cee country to l aunch a globally competitive publicly developed llm, 11 2024. URL : https://therecursive.com/bulgaria-becomes-the-first -cee-country-to-launch-a-publicly-developed-ll

  19. [27]

    S. D. . A. Authority. About sdaia, 2023. URL: https://sdaia.gov.sa/en/SDAIA/about/Pages/About.aspx

  20. [28]

    Atanasova

    T. Atanasova. Bggpt, the first bulgarian language model , is launched by insait, 01 2024. URL: https://therecursive.com/bggpt-the-first-bulgarian- language-model-is-launched/

  21. [29]

    Barbereau and L

    T. Barbereau and L. Dom. Gpt-nl: Towards a public intere st large language model, 2024. URL: https://www.hiig.de/wp-content/uploads/2024/09/Barb ereau2024-GPTNL-1.pdf

  22. [30]

    M. S. Bari, Y. Alnumay, N. A. Alzahrani, N. M. Alotaibi, H . A. Alyahya, S. AlRashed, F. A. Mirza, S. Z. Alsubaie, H. A. Alahmed, G. Alabduljabbar, R. Al khathran, Y. Almushayqih, R. Alnajim, S. Alsubaihi, M. Al Mansour, M. Alrubaian, A. Ala mmari, Z. Alawami, A. Al- Thubaity,...

  23. [31]

    Ernie bot: Baidu’s knowledge-enhance d large language model built on full ai stack technology, 2023

    Baidu Research. Ernie bot: Baidu’s knowledge-enhance d large language model built on full ai stack technology, 2023. URL: https://research.baidu.com/Blog/index-view?id=183

  24. [32]

    About, 2024

    BharatGen. About, 2024. URL: https://bharatgen.tech/about.html

  25. [33]

    URL: https://bharatgen.tech

    Bharatgen, 2024. URL: https://bharatgen.tech

  26. [34]

    D. Barry. Why tech companies will likely ignore the eu’s ai act. Reworked, 2023. URL: https://www.reworked.co/digital-workplace/why-tech- companies-will-likely-ignore-the-eus-ai-act/

  27. [35]

    S. E. D. Board. Mci, disg, sndgo, and google cloud launch ai trailblazers initiative to ac- celerate the development of impactful generative ai soluti ons in singapore, 07 2023. URL: https://www.edb.gov.sg/en/about-edb/media-releases- publications/mci-disg-sndgo-and-google-cloud-

  28. [36]

    S. P. Boggs. Japan’s new draft guidelines on ai and copyr ight: Is it really ok to train ai using pirated materials?, March 2024 . URL: https://www.privacyworld.blog/2024/03/japans-new-dr aft-guidelines-on-ai-and-copyright-is-it-reall

  29. [37]

    URL: https://bigscience.huggingface.co

    Bigscience research workshop, 2021. URL: https://bigscience.huggingface.co

  30. [38]

    G. Breda. Open government act (wet open overheid - woo). URL: https://www.breda.nl/en/open-government-act-wet-ope n-overheid-woo

  31. [39]

    H. Brown. Ai adoption could boost european business but what’s holding us back?, 07 2024. URL: https://www.euronews.com/business/2024/07/22/the-bi g-question-can-ai-really-add-600bn-to-europes 35

  32. [40]

    Brand˜ ao

    R. Brand˜ ao. The current scenario of artificial intelli gence development in brazil map- ping artificial intelligence centers in brazil: Initiative s, actions, and projects, Apr 2024. URL: https://cetic.br/media/docs/publicacoes/6/20240514085413/iso-year-xvi-n-1-ia-development-in-braz

  33. [41]

    URL: https://vlada.mk/node/32097?ln=sq

    Bytyqi: Po prezantojm¨ e asistentin e par¨ e digjital “a da”, bazuar n¨ e inteligjenc¨ en artificiale, n¨ e sektorin publik n¨ e rajon, 02 2023. URL: https://vlada.mk/node/32097?ln=sq

  34. [42]

    W. N. Caballero and P. R. Jenkins. On large language mode ls in national security applications,

  35. [43]

    URL: https://www.novinite.com/articles/229448/Bulgaria+Leads+Europe+with+First+State-Supported+AI+Langu

    Bulgaria leads europe with first state-supported ai lan guage model - novinite.com - sofia news agency, 2025. URL: https://www.novinite.com/articles/229448/Bulgaria+Leads+Europe+with+First+State-Supported+AI+Langu

  36. [44]

    potemki n ai

    M. Can. Under the leadership of our president: “potemki n ai” and the turkish approach to artificial intelligence. Third World Quarterly , page 1–21, December 2022. URL: https://www.tandfonline.com/doi/citedby/10.1080/01436597.2022.2147059?scroll=top&needAccess=true

  37. [45]

    A. P. Center. Aws to launch an infrastructure re- gion in the kingdom of saudi arabia, 03 2024. URL: https://press.aboutamazon.com/2024/3/aws-to-launch- an-infrastructure-region-in-the-kingdom-of-sau

  38. [46]

    URL: https://arxiv.org/abs/2407.03453

  39. [47]

    URL: https://indiaai.gov.in/news/cabinet-approves-india- ai-mission-at-an-outlay-of-rs-10-372-crore

    Cabinet approves india ai mission at an outlay of rs 10,3 72 crore, 2024. URL: https://indiaai.gov.in/news/cabinet-approves-india- ai-mission-at-an-outlay-of-rs-10-372-crore

  40. [48]

    URL: https://www.clarin.eu/news/clarin-eric-statement-uk raine

    Clarin eric statement on ukraine | clarin eric, 03 2022. URL: https://www.clarin.eu/news/clarin-eric-statement-uk raine

  41. [49]

    URL: https://www.clarin.eu/content/clarin-nutshell

    Clarin in a nutshell | clarin eric, 2020. URL: https://www.clarin.eu/content/clarin-nutshell

  42. [50]

    B. S. Center. Collaboration agreement between bsc and n vidia to develop solutions integrating ai and supercomputing, 2023 . URL: https://www.bsc.es/news/bsc-news/collaboration-agre ement-between-bsc-and-nvidia-develop-solution

  43. [51]

    S. Cheung. Prc adapts meta’s llama for military and secu rity ai applications, 10 2024. URL: https://jamestown.org/program/prcs-adaptation-of-op en-source-llm-for-military-and-security-purpos

  44. [52]

    Clarin-pl language resources

    CLARIN-PL. Clarin-pl language resources. URL: https://clarin-pl.eu/catalog/resources

  45. [53]

    Clova studio

    CLOV A. Clova studio. URL: https://clova.ai/en/clova-studio

  46. [54]

    About clarin-pl

    CLARIN-PL. About clarin-pl. URL: https://clarin-pl.eu/about

  47. [55]

    Clarin-pl chat

    CLARIN-PL. Clarin-pl chat. URL: https://clarin-pl.eu/tools/chat

  48. [56]

    URL: https://www.codaveri.com

    Codaveri - automated programming coach. URL: https://www.codaveri.com

  49. [57]

    Commission

    E. Commission. Digital europe programme - performance . URL: https://commission.europa.eu/strategy-and-policy/eu -budget/performance-and-reporting/programme-per

  50. [58]

    CMKL University | Thailand

    CMKL University. CMKL University | Thailand. URL: https://www.cmkl.ac.th/

  51. [59]

    N. A. Coalition. Codi, the virtual policy assistant, 06 2024. URL: https://nlaic.com/en/use_cases/codi-the-virtual-pol icy-assistant/

  52. [60]

    Commission

    E. Commission. Horizon europe, 2021

  53. [61]

    Commission

    E. Commission. The european high performance computin g joint undertaking | shaping europe’s digital future, 06 2023. URL : https://digital-strategy.ec.europa.eu/en/policies/h igh-performance-computing-joint-undertaking

  54. [62]

    Commission

    E. Commission. Digital government factsheet: The neth erlands, 2019. URL: https://interoperable-europe.ec.europa.eu/sites/def ault/files/inline-files/Digital_Government_Fa

  55. [63]

    Commission

    E. Commission. Digital programme | shaping europe’s di gital future, 2021. URL: https://digital-strategy.ec.europa.eu/en/activities /digital-programme

  56. [64]

    Commission

    E. Commission. European research development and depl oyment of ai, 11 2024

  57. [65]

    Commission

    E. Commission. New horizon europe funding boosts europ ean research in ai and quantum technologies. Shaping Europe’s Digital Future , 2024. URL: https://digital-strategy.ec.europa.eu/en/news/new-h orizon-europe-funding-boosts-european-researc

  58. [66]

    Commission

    E. Commission. Ai factories, 2024. URL: https://digital-strategy.ec.europa.eu/en/policies/a i-factories. 36

  59. [67]

    Commission

    E. Commission. Commission launches ai innovation pack age to support artificial intelligence startups and smes, 01 2024

  60. [68]

    M. . Company. Economic potential of generative ai | mcki nsey, 06 2023. URL: https://www.mckinsey.com/capabilities/mckinsey-digi tal/our-insights/the-economic-potential-of-gene

  61. [69]

    URL: https://www.ic.etf.bg.ac.rs/projects/comtext-sr/

    Comtext.sr - icef, 2018. URL: https://www.ic.etf.bg.ac.rs/projects/comtext-sr/

  62. [70]

    Commission

    E. Commission. Eu funding & tenders portal, 2025. URL: https://ec.europa.eu/info/funding-tenders/opportuni ties/portal/screen/opportunities/topic-detail

  63. [71]

    N. C. Commission. Industry statistics, 01 2025. URL: https://www.ncc.gov.ng/statistics-reports/industry- overview

  64. [72]

    URL: https://hr-xr-xtend.ffzg.unizg.hr

    Croatian extended reality extensions (hr-xr-xtend), 2024. URL: https://hr-xr-xtend.ffzg.unizg.hr

  65. [73]

    Our toll-free national hotline for agricul ture

    Crop2Cash. Our toll-free national hotline for agricul ture. URL: https://www.crop2cash.com.ng/farm-advice

  66. [74]

    F. B. Council. Brazil: The next frontier for ai data cent ers. Forbes, 2024. URL: https://www.forbes.com/councils/forbesbusinesscouncil/2024/11/14/brazil-the-next-frontier-for-ai-

  67. [75]

    Crestana

    C. Crestana. Introducing sdaia and their latest arabic llm on azure ai model catalog, 2024. URL: https://techcommunity.microsoft.com/blog/aiplatformblog/introducing-sdaia-and-their-latest-arabic

  68. [76]

    Data and A

    S. Data and A. I. A. (SDAIA). Open data policy, May 2020. U RL: https://sdaia.gov.sa/ar/SDAIA/about/Documents/Open%20Data%20Policy.pdf

  69. [77]

    Netherlands ap publishes guidance on th e entrance into effect of the eu ai act, 2024

    DataGuidance. Netherlands ap publishes guidance on th e entrance into effect of the eu ai act, 2024. URL: https://www.dataguidance.com/news/netherlands-ap-pu blishes-guidance-entrance-effect-eu

  70. [78]

    Ai services, n.d

    CyberAgent. Ai services, n.d. URL: https://www.cyberagent.co.jp/en/service/ai/

  71. [79]

    Data and A

    S. Data and A. I. A. (SDAIA). National strategy for data & ai, 2020. URL: https://ai.sa/Brochure_NSDAI_Summit%20version_EN.pdf

  72. [80]

    Secretar ´ ıa de econom ´ ıa, 2023

    Developmentaid. Secretar ´ ıa de econom ´ ıa, 2023. URL: https://www.developmentaid.org/donors/view/186044/secretaria-de-economia-mexico-ministry-of-econom

  73. [81]

    V. Dhupar. India enterprises serve over a billion local language speakers using llms built with nvidia ai, 10 2024. URL: https://blogs.nvidia.com/blog/llms-indian-languages /

  74. [82]

    What’s the most popular llm?, 2023

    Definition. What’s the most popular llm?, 2023. URL: https://www.thisisdefinition.com/insights/most-popu lar-llm

  75. [83]

    Developer

    N. Developer. Nvidia sets new generative ai performanc e and scale records in mlperf training v4.0 | nvidia technical blog, 06 2 024. URL: https://developer.nvidia.com/blog/nvidia-sets-new-g enerative-ai-performance-and-scale-records-in

  76. [84]

    S. Dunn. India’s $1.24b ai mission pays off: Nvidia pre- dicts nation will be major producer. CCN, 2024. URL: https://www.ccn.com/news/technology/indias-ai-missi on-sovereign-supercomputer-nvidia/. 37

  77. [85]

    Dynamics

    D. Dynamics. Nvidia’s israel-1 supercomputer starts o perations, 11 2023. URL: https://www.datacenterdynamics.com/en/news/nvidias- israel-1-supercomputer-starts-operations/

  78. [86]

    Netherlands organis ation for applied scientific research (tno)

    Digital Skills and Jobs Platform. Netherlands organis ation for applied scientific research (tno). URL: https://digital-skills-jobs.europa.eu/en/organisati ons/netherlands-organisation-applied-scie

  79. [87]

    D. N. B. (DNB). Organisation, 2023. URL: https://www.dnb.nl/en/about-us/organisation/

  80. [88]

    Ai commission delivers urgent roadmap to swe dish government, 2024

    Ericsson. Ai commission delivers urgent roadmap to swe dish government, 2024. URL: https://www.ericsson.com/en/news/2024/11/ai-commiss ion-delivers-urgent-roadmap-to-swedish-govern

  81. [89]

    B. Eum. Ai development in south korea - 2023, 2024. URL: https://www.turingpost.com/p/korea2023

  82. [90]

    URL: https://cdn.accesspartnership.com/wp-content/upload s/2024/01/Acce

    Economic impact report, 10 2023. URL: https://cdn.accesspartnership.com/wp-content/upload s/2024/01/Acce

  83. [91]

    Opinion 28/2024 on certain data protection aspec ts related to the processing of personal data in the context of ai models, Dec 2 024

    EDPB. Opinion 28/2024 on certain data protection aspec ts related to the processing of personal data in the context of ai models, Dec 2 024. URL: https://www.edpb.europa.eu/system/files/2024-12/edp b_opinion_202428_ai-models_en.pdf

  84. [92]

    Eu ai act reaction: Tech experts say the world ’s first ai law is ’historic but bittersweet’

    Euronews. Eu ai act reaction: Tech experts say the world ’s first ai law is ’historic but bittersweet’. Euronews, 2024. URL: https://www.euronews.com/next/2024/03/16/eu-ai-act- reaction-tech-experts-say-the-worlds-first-ai-

  85. [93]

    Ai factories

    European Commission. Ai factories. URL: https://digital-strategy.ec.europa.eu/en/policies/a i-factories

  86. [94]

    Alliance for language technologies (sg) - search for funding - euroaccess - f¨ ordermittelsuche der eurovienna eu-consulting & -mana gement gmbh, 2024

    EuroAccess. Alliance for language technologies (sg) - search for funding - euroaccess - f¨ ordermittelsuche der eurovienna eu-consulting & -mana gement gmbh, 2024. URL: https://www.euro-access.eu/en/calls/1191/Alliance-f or-Language-Technologies-SG

  87. [95]

    ’potentially disastrous’: Tech sector says the eu ai act goes too far

    Euronews. ’potentially disastrous’: Tech sector says the eu ai act goes too far. Euronews, 2023. URL: https://www.euronews.com/next/2023/12/15/potentially-disastrous-for-innovation-tech-sector-

  88. [96]

    N. C. for Artificial Intelligence and N. I. T. D. A. N. Robo tics (NCAIR). National artificial intelligence strategy, August 2024. UR L: https://ncair.nitda.gov.ng/wp-content/uploads/2024/ 08/National-AI-Strategy_01082024-copy.pdf

  89. [97]

    G. A. for Developers. Gemma - a family of lightweight, st ate-of-the art open models from google. URL: https://ai.google.dev/gemma

  90. [98]

    European approach to artificial i ntelligence

    European Commission. European approach to artificial i ntelligence. URL: https://digital-strategy.ec.europa.eu/en/policies/e uropean-approach-artificial-intelligence

  91. [99]

    Netherlands ai strategy re- port

    European Commission, NL. Netherlands ai strategy re- port. Technical report, European Commission, 2022. URL: https://ai-watch.ec.europa.eu/countries/netherlands /netherlands-ai-strategy-report_en

  92. [100]

    for Economic Co-operation and D

    O. for Economic Co-operation and D. (OECD). Argentina national ai strategy, 2021. URL: https://oecd-opsi.org/wp-content/uploads/2021/02/Ar gentina-National-AI-Strategy.pdf

  93. [101]

    P. A. for Enterprise Development (PARP). Polish agenc y for enterprise development, 2023. URL: https://en.parp.gov.pl/

  94. [102]

    for Economic Co-operation and Development

    O. for Economic Co-operation and Development. Global p artnership on artificial intelligence. URL: https://www.oecd.org/en/about/programmes/global-par tnership-on-artificial-intelligence.html

  95. [103]

    for Economic Co-operation and Development

    O. for Economic Co-operation and Development. Ukraine . URL: https://www.oecd.org/en/countries/ukraine.html

  96. [104]

    Microsoft to invest $3.2 billion in swedish a i, 2024

    Fortune. Microsoft to invest $3.2 billion in swedish a i, 2024. URL: https://fortune.com/europe/2024/06/03/microsoft-inv est-3-2-billion-swedish-ai/ . 38

  97. [105]

    fugaku-llm

    Fujitsu. Release of “fugaku-llm” – a large language mo del trained on the supercomputer “fugaku”,

  98. [106]

    D. A. for Higher Education and Science. Who can partici pate in horizon europe?, 08 2024. URL: https://ufm.dk/en/research-and-innovation/funding-p rogrammes-for-research-and-innovation/eu-and-

  99. [107]

    B. C. for Language Technology. Latxa: an open language model and evaluation suite for basque | hitz zentroa, 04 2024. URL: https://www.hitz.eus/en/node/343

  100. [108]

    Galindo, K

    L. Galindo, K. Perset, and F. Sheeka. An overview of nat ional ai strategies and policies, 2021. URL: https://www.oecd.org/en/publications/an-overview-of -national-ai-strategies-and-policies_c0514

  101. [109]

    URL: https://theinvestor.vn/genai-set-to-increase-thai-g dp-by-6-by-2030-d13359.html

    Genai set to increase thai gdp by 6% by 2030, 2024. URL: https://theinvestor.vn/genai-set-to-increase-thai-g dp-by-6-by-2030-d13359.html

  102. [110]

    URL: https://www.fujitsu.com/global/about/resources/news/press-releases/2024/0510-01.html

  103. [111]

    About fujitsu, n.d

    Fujitsu. About fujitsu, n.d. URL: https://www.fujitsu.com/global/about/

  104. [112]

    Tokyo tech, tohoku university, fuji tsu, and riken start collab- oration to develop distributed training of large language m odels, 2023

    Fujitsu Limited. Tokyo tech, tohoku university, fuji tsu, and riken start collab- oration to develop distributed training of large language m odels, 2023. URL: https://www.fujitsu.com/global/about/resources/news/press-releases/2023/0522-01.html

  105. [113]

    Government

    I. Government. Ministry of science and technology, 20 23. URL: https://www.gov.il/en/departments/ministry_of_science_and_technology/govil-landing-page

  106. [114]

    Policy for the development of ar - tificial intelligence in poland from 2020, 2020

    Government of Poland. Policy for the development of ar - tificial intelligence in poland from 2020, 2020. URL: https://wp.oecd.ai/app/uploads/2021/12/Poland_Policy_for_Artificial_Intelligence_Development_in_Po

  107. [115]

    State of ai in saudi arabia

    Global AI Summit. State of ai in saudi arabia. URL: https://globalaisummit.org/Documents/StateofAIinSaudiArabia.pdf

  108. [116]

    J. Gomes. Indonesia tech leaders team with nvidia and p artners to launch nation’s ai, 11 2024. URL: https://blogs.nvidia.com/blog/indonesia-tech-leader s-sovereign-ai/

  109. [117]

    URL: https://insait.ai/governance/

    Governance | insait, 07 2024. URL: https://insait.ai/governance/

  110. [118]

    Ai and the future of wor k, 2023

    Government of the Netherlands. Ai and the future of wor k, 2023. URL: https://open.overheid.nl/documenten/8d7b5ccb-41fc-4 d92-a4c2-a7fa81e66bfb/file

  111. [119]

    Government-wide visi on on generative ai of the netherlands, January 2024

    Government of the Netherlands. Government-wide visi on on generative ai of the netherlands, January 2024. URL: https://www.government.nl/documents/parliamentary-d ocuments/2024/01/17/government-wide-vision-on-

  112. [120]

    URL: https://www.vision2030.gov.sa/en

    Vision 2030, 2016. URL: https://www.vision2030.gov.sa/en

  113. [121]

    Ai commission has presented pro posals to the government, December

    Government of Sweden. Ai commission has presented pro posals to the government, December

  114. [122]

    URL: https://www.government.se/press-releases/2024/12/ai -commission-has-presented-proposals-

  115. [123]

    Netherlands strategi c ac- tion plan for artificial intelligence, October 2019

    Government of the Netherlands. Netherlands strategi c ac- tion plan for artificial intelligence, October 2019. URL: https://wp.oecd.ai/app/uploads/2021/12/Netherlands_Strategic_Action_Plan_for_Artificial_Intelligen

  116. [124]

    Han-ah and J

    J. Han-ah and J. Yeong-su. First steps to trust-based a i data norms, 2024. URL: https://www.pipc.go.kr/np/cop/bbs/selectBoardArticle.do?bbsId=BS074&mCode=C020010000&nttId=9083

  117. [125]

    J. Harper. Pentagon sunsets generative ai task force, launches rapid capabilities cell, 12 2024. URL: https://defensescoop.com/2024/12/11/cdao-pentagon-g enerative-ai-rapid-capabilities-cell-sunse

  118. [126]

    Excellent research and in novation are encouraged in largest-ever research and innovation bill, 2 024

    Government Offices of Sweden. Excellent research and in novation are encouraged in largest-ever research and innovation bill, 2 024. URL: https://www.government.se/press-releases/2024/12/ex cellent-research-and-innovation-are-encouraged-

  119. [127]

    Turkey, 2023

    GovInsider. Turkey, 2023. URL: https://govinsider.asia/intl-en/digital-government/ country/turkey

  120. [128]

    Grants and Funding

    E. Grants and Funding. Call for applications to develo p language foundation models, 01 2024

  121. [129]

    I. A. Hamilton. Argentina sets itself up as a mecca for a i deregulation, Jun 2024. URL: https://www.thedailyupside.com/technology/artificial-intelligence/argentina-sets-itself-up-as-a-

  122. [130]

    URL: https://aicet.comp.nus.edu.sg

    Home - ai centre for educational technologies (aicet) , 2024. URL: https://aicet.comp.nus.edu.sg

  123. [131]

    URL: https://horizon-europe.org.ua/en/heo-in-ua/

    Horizon europe office in ukraine. URL: https://horizon-europe.org.ua/en/heo-in-ua/

  124. [132]

    J. Health. About, 09 2023. URL: https://jacarandahealth.org/about/. 39

  125. [133]

    J. Health. Jacaranda launches open source llm in five african languages - jacaranda health, 08 2024. URL: https://jacarandahealth.org/jacaranda-launches-open -source-llm-in-five-african-languages/

  126. [134]

    Hoekstra

    M. Hoekstra. Transparency of ai applications in gover nment, 2023. URL: https://vector.tno.nl/en/articles/transparency-ai-a pplications-government/

  127. [135]

    Hoekstra, L

    M. Hoekstra, L. Dom, and A. F. van Veenstra. Quickscan ai in de publieke dienstverlening iii, 2025. Summary. URL: https://vector.tno.nl/publish/pages/5406/samenvatting_quickscan_ai.pdf

  128. [136]

    URL: https://www.clarin.eu/content/initiatives-supportin g-ukrainian-research-community

    Initiatives supporting the ukrainian research commu nity | clarin eric, 2022. URL: https://www.clarin.eu/content/initiatives-supportin g-ukrainian-research-community

  129. [137]

    The first five projects from inria’s partnership with the dfki | inria, 10 2020

    Inria. The first five projects from inria’s partnership with the dfki | inria, 10 2020. URL: https://www.inria.fr/en/first-five-projects-inrias- partnership-dfki

  130. [138]

    URL: https://www.clarin.hr/en/

    Hr-clarin, 2024. URL: https://www.clarin.hr/en/

  131. [139]

    Ernie model documentation, 2023

    Hugging Face. Ernie model documentation, 2023. URL: https://huggingface.co/docs/transformers/en/model_doc/ernie

  132. [140]

    G. T. Info. Ministry of industry and information techn ology, 2023. URL: https://www.globaltimes.cn/db/government/6.shtml

  133. [141]

    Sg to de - velop southeast asia’s first llm ecosystem, 2023

    Infocomm Media Development Authority (IMDA). Sg to de - velop southeast asia’s first llm ecosystem, 2023. URL: https://www.imda.gov.sg/resources/press-releases-fa ctsheets-and-speeches/press-releases/2023/sg-

  134. [142]

    G. Kang. South korea’s ai ambitions: Massive invest- ments to boost chip competitiveness. CNN, 2024. URL: https://edition.cnn.com/2024/04/09/tech/southkorea- ai-investment-chips-competitiveness-intl-hnk/in

  135. [143]

    H.-G. Kang, A. Moon, and S. Jeon. Examining the generat ive artificial in- telligence landscape: Current status and policy strategie s, March 2024. URL: https://apjis.or.kr/pdf/APJIS_34_1_150.pdf

  136. [144]

    B. E. Insight. Petrobras will invest brl 500 million in five supercomputers, 2024. URL: https://brazilenergyinsight.com/2024/08/29/petrobras-will-invest-brl-500-million-in-five-superco

  137. [145]

    URL: https://niews.nie.edu.sg/sep2022/report-card/introd ucing-artificial-intelligence-nie-ainie/

    Introducing artificial intelligence @ nie (ai@nie) | n iews #121, 2022. URL: https://niews.nie.edu.sg/sep2022/report-card/introd ucing-artificial-intelligence-nie-ainie/

  138. [146]

    Israel launches second phase of national ai program with nis 500 million investment in research and development infr astructure, 2024

    Israel Innovation Authority. Israel launches second phase of national ai program with nis 500 million investment in research and development infr astructure, 2024. URL: https://innovationisrael.org.il/en/press_release/second-phase-ai-program/

  139. [147]

    G. Kang. Eu companies warn of risks from ai law. CNN, 2023. URL: https://www.cnn.com/2023/06/30/tech/eu-companies-ri sks-ai-law-intl-hnk/index.html

  140. [148]

    South korea launches $7.5 billion push for ai manufacturing modernization

    Korea Pro. South korea launches $7.5 billion push for ai manufacturing modernization. Korea Pro , 2024. URL: https://koreapro.org/2024/09/south-korea-launches-7 -5-billion-push-for-ai-manufacturing-moderniza

  141. [149]

    Naver plans ai expansion for cross- industry partnerships through hyperclova x, 2023

    KoreaTechToday. Naver plans ai expansion for cross- industry partnerships through hyperclova x, 2023. URL: https://www.koreatechtoday.com/naver-plans-ai-expan sion-for-cross-industry-partnerships-through-

  142. [150]

    A. Kaul, S. C. Wei, N. Ridhuan, and L. Goh. Strength- ening singapore’s ai leadership with google, 05 2024. URL: https://accesspartnership.com/strengthening-singapo res-ai-leadership-with-google/

  143. [151]

    J. Kim, S. Koo, and H. Lim. Revisiting under-represent ed knowledge of latin american literature in large language models. In Proceedings of the 26th European Conference on Artificial Intelligence (ECAI 202 4), 2024. URL: https://www.researchgate.net/publication/385034101_Revis...

  144. [152]

    W. Knight. Meta and amazon voice concerns over eu ai reg ulations. CNN, 2024. URL: https://www.cnn.com/2024/05/24/tech/meta-amazon-ai- fears-eu-law/index.html

  145. [153]

    Naver all in on ai with 1 trillion won fund for ecosys- tem, plan for platform-wide integration

    Korea JoongAng Daily. Naver all in on ai with 1 trillion won fund for ecosys- tem, plan for platform-wide integration. Korea JoongAng Daily , 2024. URL: https://koreajoongangdaily.joins.com/news/2024-11-1 1/business/industry/Naver-all-in-on-AI-with-1

  146. [154]

    H. Lee, J. Yoon, B. Hwang, S. Joe, S. Min, and Y. Gwon. Kor ealbert: Pretraining a lite bert model for korean language understanding, 2021. URL: https://arxiv.org/abs/2101.11363

  147. [155]

    Leesa-nguansuk

    S. Leesa-nguansuk. Nectec, agencies roll out open tha igpt. Bangkok Post , 2023. URL: https://www.bangkokpost.com/life/tech/2556324/nectec-agencies-roll-out-open-thaigpt

  148. [156]

    Home, n.d

    Kotoba Tech. Home, n.d. URL: https://www.kotoba.tech/jp/home

  149. [157]

    Kovacevic

    G. Kovacevic. University develops ai language techno logy for serbia, 05 2024. URL: https://www.eib.org/en/stories/serbia-artificial-in telligence-belgrade-university

  150. [158]

    C. Laidley. What to expect from the magnificent seven in 2025, 2025

  151. [159]

    URL: https://www.clarin.eu/resource-families/language-mo dels

    Language models | clarin eric, 2021. URL: https://www.clarin.eu/resource-families/language-mo dels

  152. [160]

    N.-T. M. G. Ltd. Nvidia’s israel-1 ranks no. 34 in the to p500 list of the world’s most powerful supercomputers | new-techeurope, 11 2024. URL: https://www.new-techeurope.com/2024/11/24/nvidias-i srael-1-ranks-no-34-in-the-top500-list-of-the-

  153. [161]

    Saudi aramco introduces aramcome tabrain ai model and sail initiative, 2024

    Lucidity Insights. Saudi aramco introduces aramcome tabrain ai model and sail initiative, 2024. URL: https://lucidityinsights.com/news/aramco-aramcometa brain-and-sail-initiative

  154. [162]

    URL: https://www.eumonitor.eu/9353000/1/j4nvhdfcs8bljza_j9vvik7m1c3gyxp/vkp1fqrgymox

    Legal provisions of com(2018)434 - digital europe pro gramme for the period 2021-2027 - eu monitor, 2018. URL: https://www.eumonitor.eu/9353000/1/j4nvhdfcs8bljza_j9vvik7m1c3gyxp/vkp1fqrgymox

  155. [163]

    M. E. Liew. 3 ways singapore is trialling ai in educatio n, 01 2023. URL: https://govinsider.asia/intl-en/article/3-ways-Sing apore-is-trialling-AI-in-education

  156. [164]

    URL: https://ec.europa.eu/info/funding-tenders/opportuni ties/docs/2021-2027/digital/guidance/list-3rd-

    List of participating countries in the digital europe programme efta coun- tries which are part of the european economic area (eea), 10 2 024. URL: https://ec.europa.eu/info/funding-tenders/opportuni ties/docs/2021-2027/digital/guidance/list-3rd-

  157. [165]

    N. Lomas. Eu privacy body weighs in on some tricky genai lawfulness questions | techcrunch, Dec 2024. URL: https://techcrunch.com/2024/12/18/eu-privacy-body-w eighs-in-on-some-tricky-genai-lawfulness-quest

  158. [166]

    Maritaca ai: Ai brasileira com foco em llm para portuguˆ es

    Maritaca AI. Maritaca ai: Ai brasileira com foco em llm para portuguˆ es. URL: https://www.maritaca.ai/en

  159. [167]

    Sabi´ a-7b: Portuguese language model, 2 023

    Maritaca AI. Sabi´ a-7b: Portuguese language model, 2 023. URL: https://huggingface.co/maritaca-ai/sabia-7b

  160. [168]

    Luo and X

    Y. Luo and X. Dan. A practitioner’s guide to china’s new generative ai regulations, August 2023. URL: https://www.cov.com/-/media/files/corporate/publica tions/2023/08/a-practitioners-guide-to-ch

  161. [169]

    P. Mah. Alibaba cloud unveils new ai models, 2024. URL: https://www.cdotrends.com/story/4227/alibaba-cloud- unveils-new-ai-models

  162. [170]

    Mahindra

    T. Mahindra. Tech mahindra announces ai center of exce llence, powered by nvidia ai enterprise and omniverse platforms | tech mahindr a, 10 2024. URL: https://www.techmahindra.com/insights/press-release s/tech-mahindra-announces-ai-center-excellence-

  163. [171]

    C. Malin. Aramco launches world’s largest industrial llm!, 2024. URL: https://www.middleeastainews.com/p/aramco-launches- largest-industrial-llm. 41

  164. [172]

    Mendoza-Revilla, E

    J. Mendoza-Revilla, E. Trop, L. Gonzalez, M. Roller, H . Dalla-Torre, B. P. de Almeida, G. Richard, J. Caton, N. L. Carranza, M. Skwark, A. Laterre, K . Beguir, T. Pierrot, and M. Lopez. A foundational large language model for edible pla nt genomes. Communications Biology, 7(8...

  165. [173]

    Llama 3.2

    Meta. Llama 3.2. URL: https://www.llama.com

  166. [174]

    Martin, B

    L. Martin, B. Muller, P. J. O. Su´ arez, Y. Dupont, L. Rom ary, E. de la Clergerie, D. Seddah, and B. Sagot. Camembert: a tasty french language model. arXiv:1911.03894 [cs] , 05 2020. URL: https://arxiv.org/abs/1911.03894

  167. [175]

    Matthews

    D. Matthews. Ai act agreement gets mixed reaction from european tech. Science|Business, 2023. URL: https://sciencebusiness.net/news/ai/ai-act-agreemen t-gets-mixed-reaction-european-tech

  168. [176]

    McKenzie

    B. . McKenzie. Ai regulation in thailand: Current stat e and future directions, 2023. URL: https://insightplus.bakermckenzie.com/bm/attachment_dw.action?attkey=FRbANEucS95NMLRN47z%2BeeOgEFC

  169. [177]

    URL: https://www.greeknewsagenda.gr/meltemi-the-first-gr eek-large-language-model/

    Meltemi: The first greek large language model, 06 2024. URL: https://www.greeknewsagenda.gr/meltemi-the-first-gr eek-large-language-model/

  170. [178]

    Mimran and G

    T. Mimran and G. Dahan. Artificial intelligence in the b at- tlefield: A perspective from israel, Apr 2024. URL: http://opiniojuris.org/2024/04/20/artificial-intell igence-in-the-battlefield-a-perspective-from-

  171. [179]

    Isra el national ai program 2024, 2024

    Ministry of Innovation, Science and Technology. Isra el national ai program 2024, 2024. URL: https://www.gov.il/BlobFolder/news/most-news2023121 8/he/Israel%20National%20AI%20Program%202024.pd

  172. [180]

    Ada: The digital assistant based on artificial intelligence that will provide citizens with information about state aid , April 2023

    Meta.mk News Agency. Ada: The digital assistant based on artificial intelligence that will provide citizens with information about state aid , April 2023. URL: https://metamorphosis.org.mk/en/blog/ada-the-digita l-assistant-based-on-artificial-intelligence-

  173. [181]

    Tecnol´ ogico de monterrey pioneers ai-po wered learning ecosystem, 09 2024

    Microsoft. Tecnol´ ogico de monterrey pioneers ai-po wered learning ecosystem, 09 2024. URL: https://news.microsoft.com/source/latam/features/ai/tecnologico-de-monterrey-ai-ecosystem/?lang=e

  174. [182]

    URL: https://conecta.tec.mx/en/news/national/institution/microsoft-highlights-tecgpt-example-ai-mexico

    Microsoft highlights tecgpt as example of ai in mexico , 2024. URL: https://conecta.tec.mx/en/news/national/institution/microsoft-highlights-tecgpt-example-ai-mexico

  175. [183]

    Sdaia’s allam arabic llm now live o n watsonx, 2024

    Middle East AI News. Sdaia’s allam arabic llm now live o n watsonx, 2024. URL: https://www.middleeastainews.com/p/sdaias-allam-ara bic-llm-live-on-watsonx

  176. [184]

    Technology | mistral ai | frontier ai in you r hands

    Mistral AI. Technology | mistral ai | frontier ai in you r hands. URL: https://mistral.ai/technology/

  177. [185]

    Recent developments in artificial i ntelligence, 2024

    Moro˘ glu Arseven. Recent developments in artificial i ntelligence, 2024. URL: https://www.morogluarseven.com/news-and-publication s/recent-developments-in-ai

  178. [186]

    Ministry of Science and Technology

    I. Ministry of Science and Technology. Turkey: Minist ry of science, industry and technology,

  179. [187]

    URL: https://www.santander.com.mx/ceb/2024/mexico-lidera -crecimiento-de-empresas-de-intelige

    2024. URL: https://www.santander.com.mx/ceb/2024/mexico-lidera -crecimiento-de-empresas-de-intelige

  180. [188]

    Co opera¸ c˜ ao franco-brasileira ganha for¸ ca com apoio do supercomputador santos dumont, 20 24

    Minist´ erio da Ciˆ encia, Tecnologia e Inova¸ c˜ ao. Co opera¸ c˜ ao franco-brasileira ganha for¸ ca com apoio do supercomputador santos dumont, 20 24. URL: https://www.gov.br/mcti/pt-br/acompanhe-o-mcti/noti cias/2024/11/cooperacao-franco-brasileira-ganha

  181. [189]

    IA para o Bem de Todos: Pro- posta de Plano Brasileiro de Inteligˆ encia Artificial 2024- 2028, 2024

    Minist´ erio da Ciˆ encia, Tecnologia e Inova¸ c˜ ao. IA para o Bem de Todos: Pro- posta de Plano Brasileiro de Inteligˆ encia Artificial 2024- 2028, 2024. URL: https://www.gov.br/mcti/pt-br/acompanhe-o-mcti/cct/ legislacao/arquivos/IA_para_o_Bem_de_Todos.pdf

  182. [190]

    Estrat´ egia brasileira de inteligˆ encia arti- ficial (ebia)

    Minist´ erio da Ciˆ encia, Tecnologia e Inova¸ c˜ oes (MCTI). Estrat´ egia brasileira de inteligˆ encia arti- ficial (ebia). Technical report, Minist´ erio da Ciˆ encia, Tecnologia e Inova¸ c˜ oes (MCTI), 2021. URL: https://www.gov.br/mcti/pt-br/acompanhe-o-mcti/tran sformacaod...

  183. [191]

    URL: https://www.clarin.eu/content/national-consortia

    National consortia | clarin eric, 2020. URL: https://www.clarin.eu/content/national-consortia

  184. [192]

    Nstda’s supercomputer ranks 70th in the new top500 list of the world’s most powerful supercomputers, 2022

    National Science and Technology Development Agency ( NSTDA). Nstda’s supercomputer ranks 70th in the new top500 list of the world’s most powerful supercomputers, 2022. URL: https://www.nstda.or.th/en/news/news-years-2022/nst da%E2%80%99s-supercomputer-ranks-70th-in-the-

  185. [193]

    Muzoriwa

    K. Muzoriwa. Huawei cloud drives saudi arabia’s digit al transformation with ai, 09 2024. URL: https://gulfbusiness.com/how-huawei-is-fueling-saud i-arabias-digital-leap

  186. [194]

    H. X. T. NA VER Cloud. Hyperclova x technical report, 20 24. URL: https://arxiv.org/html/2404.01954v1

  187. [195]

    P. E. Nacional. Programa de transparencia y protecci´ on de datos personales en el uso de la inteligencia artificial, Sep 2023. URL: https://www.argentina.gob.ar/noticias/programa-de-t ransparencia-y-proteccion-de-datos-personales

  188. [196]

    S. R. Nambiar. Argentina personal data protection law , 2023. URL: https://www.cookieyes.com/blog/argentina-personal-d ata-protection-law/#:~:text=The%20Argentina%20

  189. [197]

    Act on the d evelopment of artificial intelligence and establishment of trust (ai basic act), 2024

    National Assembly of the Republic of Korea. Act on the d evelopment of artificial intelligence and establishment of trust (ai basic act), 2024. Enacted on D ecember 26, 2024, to take effect in January 2026. URL: https://www.cga.ct.gov/2024/ba/pdf/2024SB-00002-R01 -BA.pdf

  190. [198]

    N19 [s] clarin - case study, 2022

    Netguru. N19 [s] clarin - case study, 2022. URL: https://www.netguru.com/clients/clarin-case-study

  191. [199]

    Netherlands AI Coalition. Ained. URL: https://nlaic.com/en/partner/ained/

  192. [200]

    URL: https://www.velotix.ai/privacy-regulations/netherla nds-data-protection-act-uavg/

    Netherlands data protection act (uavg): What you need to know, Sep 2024. URL: https://www.velotix.ai/privacy-regulations/netherla nds-data-protection-act-uavg/

  193. [201]

    G. Neto, V. Costa, and W. Gaspar. Brazil’s artificial in telligence plan (pbia) of 2024: Enabler of ai sovereignty? The African Journal of Information and Communication (AJIC) , pages 1–15, 12 2024. URL: https://www.researchgate.net/publication/387496601_Brazil’s_Artificial_Int...

  194. [202]

    Naver corporation

    NA VER Corporation. Naver corporation. URL: https://www.navercorp.com/

  195. [203]

    Hyperclova x, 2023

    NA VER Corporation. Hyperclova x, 2023. URL: https://clova.ai/en/hyperclova

  196. [204]

    About naver labs europe

    NA VER LABS Europe. About naver labs europe. URL: https://europe.naverlabs.com/about/

  197. [205]

    Japan to issue grants to build academic ti es with europe on ai chips

    Nikkei Asia. Japan to issue grants to build academic ti es with europe on ai chips. Nikkei Asia , 2024. URL: https://asia.nikkei.com/Business/Education/Japan-to -issue-grants-to-build-academic-ties-with-Eur

  198. [206]

    of Brazil

    G. of Brazil. Science, technology and innovation, 202 3. URL: https://www.gov.br/mre/en/subjects/science-technolo gy-and-innovation

  199. [207]

    Netherlands forens ic institute (nfi)

    Netherlands Forensic Institute. Netherlands forens ic institute (nfi). URL: https://www.forensicinstitute.nl/

  200. [208]

    of ICT (Colombia)

    M. of ICT (Colombia). Information and communication t echnologies, 2023. URL: https://www.mintic.gov.co/portal/715/w3-propertyval ue-133096.html

  201. [209]

    G. C. News. Data protection enforcement in colombia, 2 023. URL: https://www.globalcompliancenews.com/data-privacy/d ata-protection-enforcement-in-colombia/

  202. [210]

    Newsroom

    I. Newsroom. Through partnership with ibm, saudi data and artificial intelligence author- ity (sdaia) launches a groundbreaking arabic ai model to the middle east, 05 2024. URL: https://mea.newsroom.ibm.com/sdaia-launches-allam-o n-watsonx

  203. [211]

    Cultivating company culture: The secret sauc e of successful israeli tech startups, 2023

    Newxel. Cultivating company culture: The secret sauc e of successful israeli tech startups, 2023. URL: https://newxel.com/blog/cultivating-company-culture -israeli-tech-startups/. 43

  204. [212]

    of Life Institute

    F. of Life Institute. Ai act overview, November 2024. U RL: https://artificialintelligenceact.eu/wp-content/upl oads/2024/11/Future-of-Life-InstituteAI-Act-ove

  205. [213]

    of Science and ICT

    M. of Science and ICT. About msit, 2023. URL: https://www.msit.go.kr/eng/contents/cont.do?sCode=eng&mPid=19&mId=20

  206. [214]

    of Foreign Affairs of the People’s Republic of China

    M. of Foreign Affairs of the People’s Republic of China. China’s ai capacity-building action plan for good and for all, Septemb er 2024. URL: https://www.mfa.gov.cn/eng/wjbzhd/202409/t20240927_11498465.html

  207. [215]

    of the European Union

    C. of the European Union. European high performance co mputing joint undertaking: Council adopts regulation, 07 2021

  208. [216]

    of Internal Affairs and Communications

    M. of Internal Affairs and Communications. Japan: Mini stry of internal affairs and communica- tions (mic), 2023. URL: https://globaledge.msu.edu/global-resources/resourc e/10557

  209. [217]

    of Israel

    G. of Israel. Israel national ai program 2024, Decembe r 2023. URL: https://www.gov.il/BlobFolder/news/most-news2023121 8/he/Israel%20National%20AI%20Program%202024.pd

  210. [218]

    of Japan

    G. of Japan. Ai strategy 2022, April 2022. URL: https://www8.cao.go.jp/cstp/ai/aistratagy2022en.pdf

  211. [219]

    T. G. P. on Artificial Intelligence. About - gpai, 2020. URL: https://gpai.ai/about/

  212. [220]

    URL: https://www.ai.se/en/news/open-release-first-large- nordic-language-model-gpt-sw3

    Open release of the first large nordic language model gp t-sw3 | ai sweden, 10 2023. URL: https://www.ai.se/en/news/open-release-first-large- nordic-language-model-gpt-sw3

  213. [221]

    of Thailand

    G. of Thailand. Ai thailand: National ai strategy and a ction plan (2022 – 2027), July 2022. URL: https://ai.in.th/en/about-ai-thailand/

  214. [222]

    Openthaigpt: Developing a thai chatbot system

    OpenThaiGPT. Openthaigpt: Developing a thai chatbot system. URL: https://github.com/OpenThaiGPT/openthaigpt

  215. [223]

    S. C. of the People’s Republic of China. Next genera- tion artificial intelligence development plan, July 2017. U RL: http://fi.china-embassy.gov.cn/eng/kxjs/201710/P020 210628714286134479.pdf

  216. [224]

    of the Republic of Turkey

    P. of the Republic of Turkey. National artificial intel ligence strategy 2021-2025, August 2021. URL: https://wp.oecd.ai/app/uploads/2021/12/Turkey_National_Artificial_Intelligence_Strategy_2021

  217. [225]

    Draft royal decree on artificial intelligence system service business, October 2022

    Office of the National Digital Economy and Society Commi ssion. Draft royal decree on artificial intelligence system service business, October 2022. URL: https://bact.cc/f/2022/10/202210-draft-service-busi ness-that-use-ai-regulation-hearing-chula-onde

  218. [226]

    URL: https://www.scbeic.com/en/detail/product/outlook-q3 2024

    Outlook quarter 3/2024, 09 2024. URL: https://www.scbeic.com/en/detail/product/outlook-q3 2024

  219. [227]

    S. R. Pacheco. Projeto de lei n° 2338, de 2023, 2023. URL : https://legis.senado.leg.br/sdleg-getter/documento? dm=9347622&disposition=inline

  220. [228]

    Chatgpt overview, 2025

    OpenAI. Chatgpt overview, 2025. URL: https://openai.com/chatgpt/overview/

  221. [229]

    Partnership

    A. Partnership. Strengthening singapore’s ai leader ship with google |, 05 2024. URL: https://cdn.accesspartnership.com/wp-content/upload s/2024/05/Economic-Impact-Report-Strengthening

  222. [230]

    Sponsors

    OpenThaiGPT. Sponsors. URL: https://openthaigpt.aieat.or.th/sponsors

  223. [231]

    M. Orbach. Ai21 unveils upgraded long-context jamba m odels for enterprises. Calcalist Tech,

  224. [232]

    URL: https://www.calcalistech.com/ctechnews/article/b1a4rcesc

  225. [233]

    U. O. D. Osobowych. European data protection board, 20 23. URL: https://uodo.gov.pl/en/p/european-data-protection-b oard. 44

  226. [234]

    S. E. published. Openai sneaks out website update, no longer lists microsoft as minority owner, 12 2023. URL: https://www.windowscentral.com/microsoft/openai-sne aks-out-website-update-no-longer-lists-micros

  227. [235]

    Qwen: The official repository of qwen chat & pret rained large language model proposed by alibaba cloud

    QwenLM. Qwen: The official repository of qwen chat & pret rained large language model proposed by alibaba cloud. URL: https://github.com/QwenLM/Qwen

  228. [236]

    Parliament and C

    E. Parliament and C. of the European Union. Regulation (eu) 2024/1689 laying down harmonized rules on artificial intelligence, 2024. Official Journal of th e European Union, OJ L 2024/1689, 12 July 2024 URL: https://eur-lex.europa.eu/eli/reg/2024/1689/oj/eng

  229. [237]

    D. P. Report. The edpb opinion on training ai models usi ng personal data and recent garante fine: Lawful deployment of llms, January 2 025. URL: https://www.dataprotectionreport.com/2025/01/the-ed pb-opinion-on-training-ai-models-using-person

  230. [238]

    Brazilian researchers develop their own chatg pt, 2024

    Peq42. Brazilian researchers develop their own chatg pt, 2024. URL: https://peq42.com/blog/brazilian-researchers-develo p-their-own-chatgpt/

  231. [239]

    Persoonsgegevens

    A. Persoonsgegevens. Current information, 2023. URL : https://www.autoriteitpersoonsgegevens.nl/en/current

  232. [240]

    Persoonsgegevens

    A. Persoonsgegevens. Ai act comes into effect: work to b e done for developers and users, 2024. URL: https://www.autoriteitpersoonsgegevens.nl/en/current/ai-act-comes-into-effect-work-to-be-do

  233. [241]

    D. Piper. Brazil, 2023. URL: https://www.dlapiperdataprotection.com/index.html?t=authority&c=BR

  234. [242]

    A. Roy. National strategy for artificial intelligence , 06 2018. URL: https://www.niti.gov.in/sites/default/files/2023-03 /National-Strategy-for-Artificial-Intelligence

  235. [243]

    G. SA. Open government data in saudi arabia, 2023. URL: https://www.my.gov.sa/wps/portal/snp/eParticipation/openData/?lang=en

  236. [244]

    J. Reichel. Regulating automation of swedish public a dministration, 2023. URL: https://ceridap.eu/regulating-automation-of-swedish -public-administration/?lng=en

  237. [245]

    Sahlgren

    M. Sahlgren. What is gpt-sw3? AI Sweden , 2022. URL: https://insights.ai.se/what-is-gpt-sw3

  238. [246]

    G. V. Research. Latin america large language model pow ered tools market size & outlook, 2030. Grand View Research , 2025. URL: https://www.grandviewresearch.com/horizon/outlook/large-language-model-powered-tools-market/latin

  239. [247]

    Google to invest $1 billion in thai data centr e, cloud infrastructure

    Reuters. Google to invest $1 billion in thai data centr e, cloud infrastructure. Reuters, 2024. URL: https://www.reuters.com/technology/google-invest-1- billion-thai-data-centre-cloud-infrastructure

  240. [248]

    About riken, n.d

    RIKEN. About riken, n.d. URL: https://www.riken.jp/en/about/

  241. [249]

    About rise

    RISE Research Institutes of Sweden. About rise. URL: https://www.ri.se/en/about-rise

  242. [250]

    chat gpt

    SBT News. Pesquisadores da unicamp desenvolvem “chat gpt" e "dall-e” brasileiros, 2024. URL: https://sbtnews.sbt.com.br/noticia/tecnologia/mirando-a-open-ai-pesquisadores-da-unicamp-desenvo

  243. [251]

    Science and T

    N. Science and T. Organization. Sto-activities - ist- 207, 01 2024. URL: https://www.sto.nato.int/Lists/test1/activitydetails.aspx?ID=17424

  244. [252]

    Sahabat-ai, 2017

    Sahabat-AI. Sahabat-ai, 2017. URL: https://sahabat-ai.com

  245. [253]

    Y. Serhan. How israel uses ai in gaza/emdash.cyrand what it might mean for the future of warfare, Dec

  246. [254]

    Artificial intelligence

    Samsung Research. Artificial intelligence. URL: https://research.samsung.com/artificial-intelligenc e

  247. [255]

    Under the patronage of his royal high- ness the crown prince, sdaia announces the global ai summit 2 024, 2024

    Saudi Data & AI Authority (SDAIA). Under the patronage of his royal high- ness the crown prince, sdaia announces the global ai summit 2 024, 2024. URL: https://sdaia.gov.sa/en/MediaCenter/News/Pages/NewsDetails.aspx?NewsID=296. 45

  248. [256]

    Saudi arabia’s council of ministers, 202 3

    Saudipedia. Saudi arabia’s council of ministers, 202 3. URL: https://saudipedia.com/en/article/1272/government-a nd-politics/government-affairs/saudi-arabias-

  249. [257]

    What is the open data policy in saudi arabi a?, Oct 2024

    SaudiPedia. What is the open data policy in saudi arabi a?, Oct 2024. URL: https://saudipedia.com/en/article/2779/economy-and- business/what-is-the-open-data-policy-in-saud

  250. [258]

    Spain to develop open-source llm trained in sp anish, regional languages

    Sifted. Spain to develop open-source llm trained in sp anish, regional languages. Sifted, 2024. URL: https://sifted.eu/articles/spain-large-language-mod el-generative-ai

  251. [259]

    Singapore

    A. Singapore. Sea-lion, 11 2023. URL: https://aisingapore.org/aiproducts/sea-lion/

  252. [260]

    Top websites in thailand - november 2024 most visited & popular rankings, 2024

    Semrush. Top websites in thailand - november 2024 most visited & popular rankings, 2024. URL: https://www.semrush.com/website/top/thailand/all/

  253. [261]

    S. N. Singapore. The smart nation vision. URL: https://www.smartnation.gov.sg/vision/

  254. [262]

    URL: https://time.com/7202584/gaza-ukraine-ai-warfare/

  255. [263]

    G. Shao. Digest qwen – china’s most popular open- source llm – unfold alibaba’s ai playbook, 2024. URL: https://www.aibusinessasia.com/p/china-ai-ecosystem -alibaba-qwen-innovation/

  256. [264]

    M. Sheehan. China’s ai regulations and how they get mad e, July 2023. URL: https://carnegieendowment.org/research/2023/07/chinas-ai-regulations-and-how-they-get-made?lang=

  257. [265]

    Shivanandhan

    M. Shivanandhan. Tokenizers explained – how tokenize rs help ai understand language, 2024. URL: https://www.freecodecamp.org/news/how-tokenizers-sh ape-ai-understanding/

  258. [266]

    URL: https://www.ntu.edu.sg/nie/news-events/news/detail/ shockwaves-innovations-how-nations-worldw

    Shockwaves & innovations: How nations worldwide are d ealing with ai in education, 08 2023. URL: https://www.ntu.edu.sg/nie/news-events/news/detail/ shockwaves-innovations-how-nations-worldw

  259. [267]

    Sanayi ve Teknoloji Bakanlı˘ gı

    T.C. Sanayi ve Teknoloji Bakanlı˘ gı. Ulusal yapay zek ˆ a stratejisi 2024-2025 eylem planı, 2024. URL: https://www.sanayi.gov.tr/assets/pdf/UlusalYapayZekaStratejisi2024-2025EylemPlani.pdf

  260. [268]

    E. Team. Eurollm-9b, 12 2024. URL: https://huggingface.co/blog/eurollm-team/eurollm-9b

  261. [269]

    Singapore

    A. Singapore. Southeast asian languages in one networ k data(seald), 03 2024. URL: https://aisingapore.org/aiproducts/southeast-asian- languages-in-one-network-data-seald/

  262. [270]

    Chinese ai models like alibaba’s qwen riv al openai, meta, 2024

    Tech in Asia. Chinese ai models like alibaba’s qwen riv al openai, meta, 2024. URL: https://www.techinasia.com/news/chinese-ai-models-a libabas-qwen-rival-openai-meta . 46

  263. [271]

    Smart natio n 2.0: A thriving digital future for all,

    Smart Nation and Digital Government Office. Smart natio n 2.0: A thriving digital future for all,

  264. [272]

    URL: https://file.go.gov.sg/smartnation2-report.pdf

  265. [273]

    Most used search engines by brand in south ko rea 2024, 2024

    Statista. Most used search engines by brand in south ko rea 2024, 2024. URL: https://www.statista.com/forecasts/826419/most-used -search-engines-by-brand-in-south-korea

  266. [274]

    About surf

    SURF. About surf. URL: https://www.surf.nl/en/about

  267. [275]

    URL: https://www.ai.se/en/news/swedish-government-establ ishes-new-ai-commission

    Swedish government establishes new ai commission, 20 25. URL: https://www.ai.se/en/news/swedish-government-establ ishes-new-ai-commission

  268. [276]

    A. Taylor. Albania: Pioneering ai, lagging in legal pr otections | politics | respublica, 03 2024. URL: https://respublica.edu.mk/blog-en/politics/albania- pioneering-ai-lagging-in-legal-protections

  269. [277]

    About tokyo tech, n.d

    Tokyo Institute of Technology. About tokyo tech, n.d. URL: https://www.titech.ac.jp/english/about/overview

  270. [278]

    Torres Freitas and A

    C. Torres Freitas and A. Zonaro Giacchetta. Ai watch: G lobal reg- ulatory tracker - brazil | white & case llp, May 2024. URL: https://www.whitecase.com/insight-our-thinking/ai-w atch-global-regulatory-tracker-brazil

  271. [279]

    Q. Team. About us, 2024. URL: https://qwenlm.github.io/about/

  272. [280]

    Ulusal yapay zekˆ a stratejisi 2024-2025 eylem planı yayımlandı, 2024

    T¨ urkiye Cumhuriyeti Cumhurba¸ skanlı˘ gı Dijital D¨on¨ u¸ s¨ um Ofisi. Ulusal yapay zekˆ a stratejisi 2024-2025 eylem planı yayımlandı, 2024. URL: https://cbddo.gov.tr/uyzs

  273. [281]

    URL: https://euronews.al/en/tender-opened-for-ai-project -linked-to-eu-integration/

    Tender opened for ai project linked to eu integration, 12 2023. URL: https://euronews.al/en/tender-opened-for-ai-project -linked-to-eu-integration/

  274. [282]

    URL: https://www.techmahindra.com/makers-lab/indus-proje ct

    The indus project | tech mahindra, 2024. URL: https://www.techmahindra.com/makers-lab/indus-proje ct

  275. [283]

    S. Y. Thian and Y. Hirdaramani. Singapore to channel us $52 mil- lion into building capacities for sea’s first regional llm, 2 023. URL: https://govinsider.asia/intl-id/article/singapore-t o-channel-usdollar52-million-into-building-ca

  276. [284]

    Baidu to release new version of wenxin large l anguage model in early 2025, says ceo robin li, 2024

    TMTPOST. Baidu to release new version of wenxin large l anguage model in early 2025, says ceo robin li, 2024. URL: https://en.tmtpost.com/news/7349034

  277. [285]

    Quickscan ai in public services iii, 2023

    TNO. Quickscan ai in public services iii, 2023. URL: https://www.tno.nl/en/

  278. [286]

    About tohoku university, n.d

    Tohoku University. About tohoku university, n.d. URL : https://www.tohoku.ac.jp/en/about/index.html

  279. [287]

    Eurollm-9b, 2023

    Utter Project. Eurollm-9b, 2023. URL: https://huggingface.co/utter-project/EuroLLM-9B

  280. [288]

    (VietnamPlus

    V. (VietnamPlus. Thailand prepares for national ai st rategy’s 2nd phase, 03 2024. URL: https://en.vietnamplus.vn/thailand-prepares-for-nat ional-ai-strategys-2nd-phase-post282774.vnp

  281. [289]

    Transcription and T

    U. Transcription and T. for Extended Reality. Transfo rming language translations and interpre- tations, 10 2022. URL: https://cordis.europa.eu/project/id/101070631

  282. [290]

    Next generation infrastructure for foundat ion models, 2023

    Vinnova. Next generation infrastructure for foundat ion models, 2023. URL: https://www.vinnova.se/en/p/next-generation-infrast ructure-for-foundation-models/

  283. [291]

    Udomsuvannakul, K

    P. Udomsuvannakul, K. Jirachocksubsin, S. Chueabunc hai, and P. Roongroajsat- aporn. Current stage of thailand’s regulatory approach to a i, Jul 2024. URL: https://chandler.morihamada.com/system/files/newsletters/newsletters/pdf/CMHM%20Newsletter%20-%20C

  284. [292]

    URL: https://www.clarin.eu/centres/ukrnlp-corpora

    Ukrnlp-corpora | clarin eric, 2025. URL: https://www.clarin.eu/centres/ukrnlp-corpora

  285. [293]

    About unbabel

    Unbabel. About unbabel. URL: https://unbabel.com/about/

  286. [294]

    E. H.-P. C. J. Undertaking. Winners announced for the l arge ai grand challenge, 06 2024

  287. [295]

    E. J. Undertaking. Access policy and faq, 2023. URL: https://eurohpc-ju.europa.eu/access-our-supercomput ers/access-policy-and-faq_en

  288. [296]

    E. J. Undertaking. Trustllm - large language model tec hnology for europe, 2023. URL: https://eurohpc-ju.europa.eu/trustllm-large-languag e-model-technology-europe_en

  289. [299]

    B. Vincent. Inside task force lima’s exploration of 18 0- plus generative ai use cases for dod, 11 2023. URL: https://defensescoop.com/2023/11/06/inside-task-for ce-limas-exploration-of-180-plus-generative-a

  290. [2023]

    URL: https://globaledge.msu.edu/global-resources/resourc e/10784

  291. [2024]

    URL: https://www.bunka.go.jp/english/policy/copyright/pdf/94055801_01.pdf

  292. [2025]

    URL: https://turing.iimas.unam.mx/americasnlp/

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.