Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

A Frontier AI Risk Management Framework: Bridging the Gap Between Current AI Practices and Established Risk Management

T0 review · 4 major / 6 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A frontier AI developer can keep residual risk below unacceptable levels at all times by setting an explicit risk tolerance, translating it into measurable threshold pairs, and monitoring them continuously within a governance structure.

desk verdict A competent synthesis of established risk management into a frontier-AI process framework, whose central 'ensures' claim is honestly flagged as dependent on quantitative risk-assessment methods that do not yet exist. read the letter →

arxiv 2502.06656 v3 pith:X5XNJJ5H submitted 2025-02-10 cs.AI

Simeon Campos , Henry Papadatos , Fabien Roger , Chloé Touzet , Otter Quarks , Malcolm Murray This is my paper · ORCID
classification cs.AI
keywords frontierAIriskmanagementtoleranceKeyIndicatorsControlgovernanceopen-endedredteamingassuranceprocesses
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 paper argues that frontier AI developers can manage catastrophic risks with the same formal machinery as aviation and nuclear power: set an explicit risk tolerance, translate it into measurable threshold pairs, and monitor them continuously. Its central claim is that a four-component workflow—risk identification, analysis and evaluation, treatment, and governance—can keep residual risk below unacceptable levels at all times if developers commit to it. The paper is a bridge proposal that maps current AI safety policies onto established risk-management standards and fills the gaps it finds, most notably the absence of a quantified risk tolerance. It is candid that the quantitative methods needed to define and enforce those thresholds do not yet exist, so the framework is offered as a target structure for the field to grow into. A sympathetic reader would care because the workflow is concrete enough to be adopted before regulation arrives and relocates most risk work to the planning phase, before the costly final training run.

What carries the argument

The load-bearing mechanism is the KRI/KCI threshold pair and the 'if-then' logic linking them to a risk tolerance. A Key Risk Indicator is a measurable proxy for a risk, such as success on a cybersecurity benchmark; a Key Control Indicator is a measurable proxy for the effectiveness of a mitigation, such as a containment security level. The framework asserts a three-way relationship: for any risk tolerance and KRI threshold there is a minimum KCI threshold that must be met, so setting any two of the three determines the third. Risk models—scenario-by-scenario pathways from model capabilities to real-world harms—are what make this relationship quantitative, and continuous monitoring of both indicators during training and deployment is what enforces it.

What would settle it

Run the framework on a real deployed frontier model: fix a numerical risk tolerance (for example, less than a 1% annual chance of $500 million in economic damage), derive KRI/KCI threshold pairs from a documented risk model, and then compare observed incident frequency and severity against the tolerance for a year while all KCI thresholds are met. If the observed risk exceeds the tolerance in a case where the stated KCI thresholds were satisfied, the core guarantee—that meeting KCI thresholds keeps risk below tolerance—would be falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that a frontier AI developer can make safety concrete by defining an aggregate risk tolerance—ideally as probability times severity per unit of time, or as a quantitative probability bound on a described harmful scenario—and then translating that tolerance into pairs of Key Risk Indicators (KRIs) and Key Control Indicators (KCIs) joined by if-then logic: if a KRI threshold is crossed, the corresponding KCI threshold must be met to keep residual risk below the tolerance. Risk models, built from literature taxonomies, open-ended red teaming, and probabilistic scenario analysis, supply the quantitative relationship among the three quantities, so setting any two determines the third. The framework places this threshold machinery inside a governance structure with risk owners, a chief risk officer, board-level oversight, independent audit, and transparency, and schedules most of the analytical work during the planning phase, using scaling laws to predict which thresholds will be crossed and which mitigations must be ready. The paper presents this as the rigor missing from current AI safety frameworks, which it says lack explicit risk tolerance, quantitative assessment, and systematic risk identification.

Load-bearing premise

The framework assumes that AI risks can be measured and turned into numbers well enough to set a safety limit and trigger thresholds that keep actual harm below that limit, and current methods for doing this do not yet exist.

Editorial extensions

If this is right

  • AI developers adopting the framework would publish a numeric risk tolerance before training, making their implicit safety trade-offs legible to regulators and the public.
  • Most risk-management work—risk modeling, threshold definition, and mitigation planning—would shift to the pre-training planning phase, reducing pressure to cut corners at deployment.
  • KRI/KCI triggers would give labs a concrete go/no-go rule: if a capability threshold is crossed without the required control threshold, training or deployment stops until the control is in place.
  • The framework implies that containment, deployment filters, and safety fine-tuning are insufficient once models reach high dangerous capabilities, and that assurance processes supplying affirmative safety evidence become necessary, even though none exist yet.
  • The governance component would make board-level oversight and independent audit standard practice for frontier AI firms, mirroring listed-company requirements.

Reading between the lines

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

  • The paper leaves implicit that the same if-then logic could be piloted on a low-stakes capability, such as a cyber-benchmark score with a fictional risk budget, to calibrate the three-way relationship against real incident data before it is used for catastrophic risks.
  • If the framework works, regulators could avoid prescribing specific mitigations and instead require labs to publish risk tolerances and the KRI/KCI pairs derived from them, with independent audit of whether thresholds are met; the paper gestures at this but does not specify the audit standard.
  • The paper's own limitation points to a precondition: quantitative risk assessment and assurance processes must mature for the framework to function as advertised, so those research programs, not the framework itself, are what currently stand between the proposal and its guarantee.
  • A natural extension would push the framework beyond model developers to compute providers and cloud infrastructure, since containment KCIs depend on securing weight storage and training systems that are often operated by third parties.
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

4 major / 6 minor

Summary. The paper proposes a frontier AI risk management framework with four components: risk identification, risk analysis and evaluation, risk treatment, and risk governance. It draws on established risk management practices from aviation, nuclear power, and enterprise risk management, and adapts them to the life-cycle of frontier AI development. The central mechanism is the operationalization of a quantitative risk tolerance into paired Key Risk Indicator (KRI) and Key Control Indicator (KCI) thresholds, linked by risk models so that crossing a KRI threshold triggers mandatory mitigation to meet a KCI threshold, thereby keeping residual risk below the stated tolerance. The paper also details governance structures, a risk register, and a planned sequence of risk-management activities across the planning, training, and post-deployment phases. It explicitly acknowledges in Section 5 that quantitative AI risk assessment methods are currently insufficient to rigorously demonstrate that KCI thresholds maintain risks below the risk tolerance.

Significance. If the framework were fully operational, it would give frontier AI developers a concrete, auditable process for setting risk tolerances, defining measurable triggers and mitigation targets, and assigning governance accountability before and during model training. The paper is a useful synthesis of existing AI safety frameworks (Anthropic RSP, OpenAI Preparedness, Google DeepMind FSF) with mature risk-management standards, and it makes a credible case that explicit, quantitative risk tolerance is a missing element in current practice. A strength is that the paper is careful to distinguish aspirational claims from current capabilities, and it names concrete boundary conditions, such as the nonexistence of assurance processes. Its main contribution is conceptual and organizational rather than empirical; it does not provide a worked implementation, dataset, or case study, and the core mechanism depends on quantitative risk models that the paper concedes are not yet mature. The paper would be a valuable reference for AI governance practitioners if its central guarantee is reframed as conditional on the development of those quantitative methods.

major comments (4)
  1. [Executive Summary and Section 3.2.2] The paper's central claim that following the workflow 'ensures that risks remain below unacceptable levels at all times' is not currently supported by the framework's own premises. Section 3.2.2 states that risk models determine the three-way relationship among risk tolerance, KRI thresholds, and KCI thresholds, but Section 5 concedes that 'current quantitative risk assessment methods are currently insufficient to rigorously demonstrate that KCI thresholds maintain risks below the risk tolerance.' Because the 'ensures' guarantee depends on the ability to quantify scenario-step probabilities and severities (Section 3.1.3), the guarantee is aspirational rather than operational. The paper should explicitly reframe the 'ensures' language as conditional on the maturity of quantitative AI risk assessment, or provide a concrete worked example showing how the relationship can be discharged with current methods.
  2. [Section 3.2.2, footnote 5] The illustrative example ('if a model reaches 60% on Cybench... then maintaining cyber security level 3... is required...') is presented as a template, but the footnote immediately states that establishing such quantitative links remains challenging. This is not an internal contradiction, but it means the example is purely fictional and does not demonstrate feasibility. The manuscript should either specify what evidence would be needed to validate such a link, or clearly label the example as a placeholder that presupposes the existence of risk models that have not yet been built.
  3. [Section 5] The limitations paragraph identifies exactly the load-bearing gap: the field lacks detailed understanding of how harms materialize, and quantitative assessment is insufficient. However, the limitations are stated after the framework's normative requirements have been presented as rigid obligations (e.g., 'development must be put on hold' if a KCI threshold cannot be met). This creates a mismatch between the framework's regulatory-style language and its current epistemic basis. The paper should integrate these limitations into the statement of the framework's requirements, for example by adding a 'maturity conditions' subsection that distinguishes which parts of the framework are ready for adoption today and which are contingent on future methods.
  4. [Section 2.1] The claim that existing AI safety frameworks 'deviate significantly from risk management norms, without clear justification' is supported primarily by a citation to SaferAI (2024), an assessment produced by the authors' own organization. This is a mild self-referential evidence source. To strengthen the argument, the paper should either include an independent analysis or clearly disclose the potential conflict of interest at the point of citation, rather than only in the author affiliations.
minor comments (6)
  1. [Section 3.2.2] The phrase 'this quantitative estimation could be acheived' contains a typo: 'acheived' should be 'achieved'.
  2. [Section 3.2.1] The paragraph on regulatory oversight says 'no AI developers explicitly set their risk tolerance' but later in the same paragraph says 'AI developers implicitly define their risk tolerance.' This is consistent, but the contrast could be made crisper by defining 'explicitly' as 'in a documented, legible form' at first use.
  3. [Section 4.2.2] The training-phase description says open-ended red teaming is used 'to identify any unexpected risks or emerging capabilities,' but Section 3.1.2 already defined open-ended red teaming as focused on unforeseen risks. The redundancy is fine, but the text could clarify whether capability emergence is a separate activity or part of red teaming.
  4. [Glossary] The glossary defines 'risk tolerance' as 'the aggregate level of risk that society or AI developers is willing to accept.' The singular verb 'is' should be 'are,' and the definition would benefit from distinguishing societal risk tolerance from organizational risk tolerance, since the paper explicitly says regulators should set the former.
  5. [References] Some references use inconsistent date formats (e.g., 'n.d.' for West, '2023a' and '2023b' for ISO/IEC but the in-text citations use ISO/IEC, 2009 and NIST, 2024). The reference list would benefit from a consistency pass.
  6. [Figure 2] Figure 2 is referenced as showing the complete framework, but the figure is not included in the provided text. If the figure is present in the actual manuscript, it should be checked for readability at print size, especially the examples in each component box.

Circularity Check

0 steps flagged · score 0.0 of 10

Framework is a normative synthesis, not a derivation; no circular step reduces its claims to its inputs.

full rationale

The paper does not derive empirical predictions from fitted parameters. Its central mechanism—the three-way KRI/KCI/risk-tolerance relationship in Section 3.2.2—is definitional in the sense that thresholds are to be chosen via risk models, but the paper explicitly labels the worked example as fictional and concedes in Section 5 and footnote 5 that current quantitative methods are insufficient to discharge the guarantee. That is an acknowledged gap in operational support, not a circular reduction. The framework is explicitly mapped onto external standards (Raz & Hillson, NIST AI RMF, ISO/IEC), and its governance and treatment components are grounded in independent literature. The only self-referential element is the citation to SaferAI (2024) in Section 2.1 to support the claim that existing company frameworks lack quantitative rigor; that claim is corroborated by an independent source (Institute for AI Policy and Strategy, 2024) and is motivational rather than load-bearing for the framework's internal logic. No equation or predictive claim reduces by construction to an input, so no circularity is present.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No numeric parameters are fitted anywhere in the paper; the framework is conceptual. The load-bearing assumptions are domain assumptions about quantifiability, predictability, and the future availability of assurance processes. These are stated or acknowledged in the paper, but they are not validated.

assumptions (4)
  • domain assumption Risk tolerance can be expressed as probability times severity per unit time and should be set before development.
    Section 3.2.1 imports this norm from aviation and nuclear regulation (FAA AC 25.1309, NRC safety goals) without evidence that AI risk can be quantified on the same scale.
  • domain assumption Risk modeling and expert elicitation can produce quantitative estimates for each step of an AI risk pathway.
    Section 3.1.3 and Section 3.2.2 propose event trees, Delphi surveys, and scenario probability estimates; the paper notes in Section 5 that the field lacks the detailed understanding needed for this.
  • domain assumption Scaling laws can predict capability thresholds in time to prepare mitigations before the final training run.
    Section 4.2.1 relies on scaling laws to plan ahead and acknowledges empirical evaluation is still required; prediction error would shift the risk timeline.
  • domain assumption Assurance processes that provide affirmative safety evidence for models with dangerous capabilities are achievable in principle.
    Section 3.2.2 and Section 3.3.1 require such processes for high-capability models, while the paper states that as of early 2025 none have been demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Frontier AI Risk Management Framework: Bridging the Gap Between Current AI Practices and Established Risk Management." pith.science (2026). https://pith.science/paper/X5XNJJ5H

@misc{pith2026250206656,
  author       = {Pith},
  title        = {Pith review of: A Frontier AI Risk Management Framework: Bridging the Gap Between Current AI Practices and Established Risk Management},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X5XNJJ5H}},
  note         = {Machine review of arXiv:2502.06656}
}
read the original abstract

The recent development of powerful AI systems has highlighted the need for robust risk management frameworks in the AI industry. Although companies have begun to implement safety frameworks, current approaches often lack the systematic rigor found in other high-risk industries. This paper presents a comprehensive risk management framework for the development of frontier AI that bridges this gap by integrating established risk management principles with emerging AI-specific practices. The framework consists of four key components: (1) risk identification (through literature review, open-ended red-teaming, and risk modeling), (2) risk analysis and evaluation using quantitative metrics and clearly defined thresholds, (3) risk treatment through mitigation measures such as containment, deployment controls, and assurance processes, and (4) risk governance establishing clear organizational structures and accountability. Drawing from best practices in mature industries such as aviation or nuclear power, while accounting for AI's unique challenges, this framework provides AI developers with actionable guidelines for implementing robust risk management. The paper details how each component should be implemented throughout the life-cycle of the AI system - from planning through deployment - and emphasizes the importance and feasibility of conducting risk management work prior to the final training run to minimize the burden associated with it.

Figures

Figures reproduced from arXiv: 2502.06656 by the authors.

Figure 1
Figure 1. Key components of the frontier AI risk management framework. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Our complete risk management framework along with examples illustrating each compo [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Systematic Hazard Analysis for Frontier AI using STPA

    cs.CY 2025-06 conditional novelty 5.0 of 10

    Applying STPA to the AI Control scenario produces structured unsafe control actions and loss scenarios, supporting an argument that systematic hazard analysis can improve frontier AI safety assurance.

Reference graph

Works this paper leans on

59 extracted references · 45 canonical work pages · cited by 1 Pith paper

  1. [1]

    Responsible scaling policy

    Anthropic (October 15, 2024). Responsible scaling policy. https://assets.anthropic.com/m/24a47b00f10301cd/original/Anthropic-Responsible-Scaling-Policy-2024-10-15.pdf

    work page 2024

  2. [2]

    The long-term benefit trust

    Anthropic (September 19, 2023). The long-term benefit trust. https://www.anthropic.com/news/the-long-term-benefit-trust

    work page 2023

  3. [3]

    Hobbhahn, D

    Balesni, M., M. Hobbhahn, D. Lindner, A. Meinke, T. Korbak, J. Clymer, B. Shlegeris, J. Scheurer, C. Stix, R. Shah, N. Goldowsky-Dill, D. Braun, B. Chughtai, O. Evans, D. Kokotajlo, and L. Bushnaq (2024). Towards evaluations-based safety cases for ai scheming. https://arxiv.org/abs/2411.03336

    arXiv 2024

  4. [4]

    Barrett, A. M., J. Newman, B. Nonnecke, D. Hendrycks, E. R. Murphy, and K. Jackson (2023). Ai risk-management standards profile for general-purpose ai systems (gpais) and foundation models. https://cltc.berkeley.edu/wp-content/uploads/2023/11/Berkeley-GPAIS-Foundation-Model-Risk-Management-Standards-Profile-v1.0.pdf

    work page 2023

  5. [5]

    Statement on ai risk

    Center for AI Safety (2023). Statement on ai risk. https://www.safe.ai/work/statement-on-ai-risk

    work page 2023

  6. [6]

    Gabrieli, D

    Clymer, J., N. Gabrieli, D. Krueger, and T. Larsen (2024). Safety cases: How to justify the safety of advanced ai systems. https://arxiv.org/abs/2403.10462

    arXiv 2024

  7. [7]

    Coates, J. C. (2007). The goals and promise of the sarbanes–oxley act. The Journal of Economic Perspectives\/ 21\/ (1), 91--116

    work page 2007

  8. [8]

    Coccia, M. (2018). The fishbone diagram to identify, systematize and analyze the sources of general purpose technologies. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3100011

    work page 2018

Show all 59 references
  1. [9]

    Enterprise risk management integrating with strategy and performance

    Committee of Sponsoring Organizations of the Treadway Commission (2017). Enterprise risk management integrating with strategy and performance. https://aaahq.org/portals/0/documents/coso/coso_erm_2017_-_exec_summary.pdf

    2017

  2. [10]

    Skalse, Y

    Dalrymple, D., J. Skalse, Y. Bengio, S. Russell, M. Tegmark, S. Seshia, S. Omohundro, C. Szegedy, B. Goldhaber, N. Ammann, A. Abate, J. Halpern, C. Barrett, D. Zhao, T. Zhi-Xuan, J. Wing, and J. Tenenbaum (2024). Towards guaranteed safe ai: A framework for ensuring robust and ...

    2024 arXiv

  3. [11]

    Emerging processes for frontier ai safety, ai safety summit

    Department for Science, Innovation and Technology (2023). Emerging processes for frontier ai safety, ai safety summit. https://assets.publishing.service.gov.uk/media/653aabbd80884d000df71bdc/emerging-processes-frontier-ai-safety.pdf

    2023

  4. [12]

    Frontier ai safety commitments, ai seoul summit 2024

    Department for Science, Innovation and Technology (2024). Frontier ai safety commitments, ai seoul summit 2024. https://www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024/frontier-ai-safety-commitments-ai-seoul-summit-2024

    2024

  5. [13]

    Eisenhardt, K. (1989). Making fast strategic decisions in high-velocity environments. The Academy of Management Journal\/

    1989

  6. [14]

    Strong risk culture — sound banks

    European Central Bank (2023). Strong risk culture — sound banks. https://www.bankingsupervision.europa.eu/press/publications/newsletter/2023/html/ssm.nl230215_3.en.html

    2023

  7. [15]

    Schwering, and S

    Ewelt-Knauer, C., A. Schwering, and S. Winkelmann (2020). Doing good by doing bad: How tone at the top and tone at the bottom impact performance-improving noncompliant behavior. Journal of Business Ethics\/ 175\/ (3), 609--624

    2020

  8. [16]

    Bindu, A

    Fang, R., R. Bindu, A. Gupta, Q. Zhan, and D. Kang (2024). Llm agents can autonomously hack websites. https://arxiv.org/abs/2402.06664

    2024 arXiv

  9. [17]

    System design and analysis

    Federal Aviation Administration (1988). System design and analysis. https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC_25.1309-1A.pdf

    1988

  10. [18]

    Frontier model forum: Advancing frontier ai safety and security

    Frontier Model Forum (2023). Frontier model forum: Advancing frontier ai safety and security. https://www.frontiermodelforum.org/

    2023

  11. [19]

    Goemans, A., M. D. Buhl, J. Schuett, T. Korbak, J. Wang, B. Hilton, and G. Irving (2024). Safety case template for frontier ai: A cyber inability argument. https://arxiv.org/abs/2411.08088

    2024 arXiv

  12. [20]

    Frontier safety framework

    Google DeepMind (2024). Frontier safety framework. https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/introducing-the-frontier-safety-framework/fsf-technical-report.pdf

    2024

  13. [21]

    Shlegeris, K

    Greenblatt, R., B. Shlegeris, K. Sachan, and F. Roger (2024). Ai control: Improving safety despite intentional subversion. https://arxiv.org/abs/2312.06942

    2024 arXiv

  14. [22]

    Keeney, and H

    Hammond, J., R. Keeney, and H. Raiffa (1998). The hidden traps in decision making. Harvard Business Review\/

    1998

  15. [23]

    Borgeaud, A

    Hoffmann, J., S. Borgeaud, A. Mensch, E. Buchatskaya, T. Cai, E. Rutherford, D. de Las Casas, L. A. Hendricks, J. Welbl, A. Clark, T. Hennigan, E. Noland, K. Millican, G. van den Driessche, B. Damoc, A. Guy, S. Osindero, K. Simonyan, E. Elsen, J. W. Rae, O. Vinyals, and L. Sif...

    2022 arXiv

  16. [24]

    Hsu, C.-C. and B. A. Sandford (2007). The delphi technique: making sense of consensus. Practical assessment, research, and evaluation\/ 12\/ (1)

    2007

  17. [25]

    Responsible scaling: Comparing government guidance and company policy

    Institute for AI Policy and Strategy (2024). Responsible scaling: Comparing government guidance and company policy. https://static1.squarespace.com/static/64edf8e7f2b10d716b5ba0e1/t/65f19a4a32c41d331ec54b87/1710332491804/Responsible+Scaling_+Comparing+Government+Guidance+and+C...

    2024

  18. [26]

    Development and application of level 1 probabilistic safety assessment for nuclear power plants

    International Atomic Energy Agency (2010). Development and application of level 1 probabilistic safety assessment for nuclear power plants. https://www.iaea.org/publications/8235

    2010

  19. [27]

    Iec/iso 31010:2009(en), risk management — risk assessment techniques

    ISO/IEC (2009). Iec/iso 31010:2009(en), risk management — risk assessment techniques. https://cdn.standards.iteh.ai/samples/14241/a9e056f913f74bd39bf829992d86d925/IEC-ISO-31010-2009.pdf

    2009

  20. [28]

    Iso/iec 23894:2023(fr), information technology — artificial intelligence — guidance on risk management

    ISO/IEC (2023a). Iso/iec 23894:2023(fr), information technology — artificial intelligence — guidance on risk management. https://www.iso.org/obp/ui#iso:std:iso-iec:23894:ed-1:v1:en:sec:6

    2023

  21. [29]

    Iso/iec 42001:2023(en), information technology — artificial intelligence — management system

    ISO/IEC (2023b). Iso/iec 42001:2023(en), information technology — artificial intelligence — management system. https://www.iso.org/obp/ui/en/#iso:std:iso-iec:42001:ed-1:v1:en

    2023

  22. [30]

    McCandlish, T

    Kaplan, J., S. McCandlish, T. Henighan, T. B. Brown, B. Chess, R. Child, S. Gray, A. Radford, J. Wu, and D. Amodei (2020). Scaling laws for neural language models. https://arxiv.org/abs/2001.08361

    2020 arXiv

  23. [31]

    Karnofsky, H. (2024). If-then commitments for ai risk reduction. https://carnegie-production-assets.s3.amazonaws.com/static/files/Karnofsky

    2024

  24. [32]

    Koessler, L. and J. Schuett (2023). Risk assessment at agi companies: A review of popular risk assessment techniques from other safety-critical industries. https://arxiv.org/abs/2307.08823

    2023 arXiv

  25. [33]

    Clymer, B

    Korbak, T., J. Clymer, B. Hilton, B. Shlegeris, and G. Irving (2025). A sketch of an ai control safety case. https://arxiv.org/abs/2501.17315

    2025 arXiv

  26. [34]

    Wilkinson, C

    Leveson, N., C. Wilkinson, C. Fleming, J. Thomas, and I. Tracy (2014). A comparison of stpa and the arp 4761 safety assessment process. http://sunnyday.mit.edu/STAMP/ARP4761-Comparison-Report-final-2.pdf

    2014

  27. [35]

    Lundqvist, S. A. (2015). Why firms implement risk governance – stepping beyond traditional risk management to enterprise risk management. Journal of Accounting and Public Policy\/ 34\/ (5), 441--466

    2015

  28. [36]

    Lahav, A

    Nevo, S., D. Lahav, A. Karpur, Y. Bar-On, H. A. Bradley, and J. Alstott (2024). Securing ai model weights: Preventing theft and misuse of frontier models. https://www.rand.org/pubs/research_reports/RRA2849-1.html

    2024

  29. [37]

    Confusion over risk criteria

    Nicholls and Smith (2020). Confusion over risk criteria. https://www.icheme.org/media/25697/hazards-30-paper-31-nicholls.pdf

    2020

  30. [38]

    Artificial intelligence risk management framework (ai rmf 1.0)

    NIST (2023). Artificial intelligence risk management framework (ai rmf 1.0). https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.100-1.pdf

    2023

  31. [39]

    Risk mitigation - glossary

    NIST (2024). Risk mitigation - glossary. https://csrc.nist.gov/glossary/term/risk_mitigation. Accessed: December 13, 2024

    2024

  32. [40]

    Safety goals for nuclear power plant operation

    Nuclear Regulatory Comission (1983). Safety goals for nuclear power plant operation. https://www.nrc.gov/docs/ML0717/ML071770230.pdf#page=22

    1983

  33. [41]

    Nyse regulation

    NYSE (2024). Nyse regulation. https://www.nyse.com/regulation/nyse

    2024

  34. [42]

    Guidelines for mnes - organisation for economic co-operation and development

    OECD (n.d.). Guidelines for mnes - organisation for economic co-operation and development. https://mneguidelines.oecd.org/due-diligence-guidance-for-responsible-business-conduct.htm

  35. [43]

    Preparedness framework (beta)

    OpenAI (2023). Preparedness framework (beta). https://cdn.openai.com/openai-preparedness-framework-beta.pdf

    2023

  36. [44]

    Gpt-4o system card

    OpenAI (2024). Gpt-4o system card. https://cdn.openai.com/gpt-4o-system-card.pdf#page=3.08

    2024

  37. [45]

    Bloomfield, A

    Pannu, J., D. Bloomfield, A. Zhu, R. MacKnight, G. Gomes, A. Cicero, and T. Inglesby (2024). Prioritizing high-consequence biological capabilities in evaluations of artificial intelligence models. http://dx.doi.org/10.2139/ssrn.4873106

    2024 doi

  38. [46]

    Parker, S. (2014). Just culture. https://www.caa.co.uk/media/sf3eiszu/fwm20160629_06_just-culture.pdf

    2014

  39. [47]

    Raz, T. and D. Hillson (2005). A comparative review of risk management standards. Risk Management\/ 7\/ (4), 53--66

    2005

  40. [48]

    Ruan, Y., C. J. Maddison, and T. Hashimoto (2024). Observational scaling laws and the predictability of language model performance. https://arxiv.org/abs/2405.10938

    2024 arXiv

  41. [49]

    A brief assessment of openai’s preparedness framework and some suggestions for improvement

    SaferAI (2024). A brief assessment of openai’s preparedness framework and some suggestions for improvement. https://assets-global.website-files.com/64332a76ab91bba8239ac2e0/65aab66bcf70ab8c986f67df_A

    2024

  42. [50]

    Slattery, et al. (2024). The ai risk repository: A comprehensive meta-review, database, and taxonomy of risks from artificial intelligence. https://cdn.prod.website-files.com/669550d38372f33552d2516e/66bc918b580467717e194940_The

    2024

  43. [51]

    Guidance on good practices in corporate governance disclosure

    UN Trade and Development (2006). Guidance on good practices in corporate governance disclosure. https://unctad.org/system/files/official-document/iteteb20063_en.pdf

    2006

  44. [52]

    Securities and Exchange Commission (2002)

    U.S. Securities and Exchange Commission (2002). All about auditors: What investors need to know. https://www.sec.gov/about/reports-publications/investorpubsaboutauditorshtm

    2002

  45. [53]

    Securities and Exchange Commission (2024)

    U.S. Securities and Exchange Commission (2024). Exchange act reporting and registration. https://www.sec.gov/resources-small-businesses/going-public/exchange-act-reporting-registration

    2024

  46. [54]

    Variengien, A

    Wang, K., A. Variengien, A. Conmy, B. Shlegeris, and J. Steinhardt (2022). Interpretability in the wild: a circuit for indirect object identification in gpt-2 small. https://arxiv.org/abs/2211.00593

    2022 arXiv

  47. [55]

    Uesato, M

    Weidinger, L., J. Uesato, M. Rauh, C. Griffin, P. Huang, J. Mellor, A. Glaese, M. Cheng, B. Balle, A. Kasirzadeh, C. Biles, S. Brown, Z. Kenton, W. Hawkins, T. Stepleton, A. Birhane, L. A. Hendricks, L. Rimell, W. Isaac, and I. Gabriel (2022). Taxonomy of risks posed by langua...

    2022

  48. [56]

    West, H. (n.d.). Speak up culture — what is it, and why is it important? https://www.goodcourse.co/post/what-is-speak-up-culture

  49. [57]

    Key considerations for updating 2023 annual report risk factors

    White & Case LLP (2023). Key considerations for updating 2023 annual report risk factors. https://www.whitecase.com/insight-alert/key-considerations-updating-2023-annual-report-risk-factors

    2023

  50. [58]

    Zhang, A. K., N. Perry, R. Dulepet, J. Ji, C. Menders, J. W. Lin, E. Jones, G. Hussein, S. Liu, D. Jasper, P. Peetathawatchai, A. Glenn, V. Sivashankar, D. Zamoshchin, L. Glikbarg, D. Askaryar, M. Yang, T. Zhang, R. Alluri, N. Tran, R. Sangpisit, P. Yiorkadjis, K. Osele, G. Ra...

    2024 arXiv

  51. [59]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.