Browse AI risks

2,500 entries · page 50 of 50

1. 70.03.03a · Additional evidence

   Economic Risks

   Power concentration

   —

   From Embodied AI: Emerging Risks and Opportunities for Policy Action (Perlo2025)

2. 70.04.00 · Risk Category

   Social Risks

   -

   From Embodied AI: Emerging Risks and Opportunities for Policy Action (Perlo2025)

   Not coded Not coded Not coded

3. 70.04.01a · Additional evidence

   Social Risks

   Bias and discrimination

   —

   From Embodied AI: Emerging Risks and Opportunities for Policy Action (Perlo2025)

4. 70.04.02a · Additional evidence

   Social Risks

   Lack of accountability and liability

   —

   From Embodied AI: Emerging Risks and Opportunities for Policy Action (Perlo2025)

5. 70.04.05a · Additional evidence

   Social Risks

   Transformative effects

   —

   From Embodied AI: Emerging Risks and Opportunities for Policy Action (Perlo2025)

6. 71.01.00 · Risk Category

   Scientific Domain of Agents

   -

   From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

   Not coded Not coded Not coded

7. 71.01.01a · Additional evidence

   Scientific Domain of Agents

   Chemical Risks

   —

   From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

8. 71.01.02a · Additional evidence

   Scientific Domain of Agents

   Biological Risks

   —

   From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

9. 71.01.03a · Additional evidence

   Scientific Domain of Agents

   Radiological Risks

   —

   From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

10. 71.01.04a · Additional evidence

    Scientific Domain of Agents

    Physical (Mechanical ) Risks

    —

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

11. 71.01.05a · Additional evidence

    Scientific Domain of Agents

    Information Science Risks

    —

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

12. 71.01.06 · Risk Sub-Category

    Scientific Domain of Agents

    Emerging Tech

    "Uncontrolled AI self- improvement; Quantum security"

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

    Not coded Not coded Not coded

13. 71.02.00 · Risk Category

    User Intent

    "Whether the risk originates from malicious intent or is an unintended consequence of legitimate task objectives"

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

    Not coded Not coded Not coded

14. 71.02.03 · Risk Sub-Category

    User Intent

    Unintended Consequences

    "Unpredictable and unforeseen outcomes from purposeful actions"

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

    Other Unintentional Other

15. 71.02.03a · Additional evidence

    User Intent

    Unintended Consequences

    —

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

16. 71.03.00 · Risk Category

    Environment

    -

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

    Not coded Not coded Not coded

17. 71.03.01a · Additional evidence

    Environment

    Nature

    —

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

18. 71.03.02 · Risk Sub-Category

    Environment

    Human Health

    "Damage to individual well-being or public health"

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

    Not coded Not coded Not coded

19. 71.03.02a · Additional evidence

    Environment

    Human Health

    —

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

20. 71.03.03 · Risk Sub-Category

    Environment

    Socioeconomics

    "Dramatically change the social and economic status"

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

    Not coded Not coded Not coded

21. 71.03.03a · Additional evidence

    Environment

    Socioeconomics

    —

    From Risks of AI Scientists: Prioritizing Safeguarding Over Autonomy (Tang2025)

22. 72.01.00a · Additional evidence

    Misuse Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

23. 72.01.02a · Additional evidence

    Misuse Risks

    Biological and Chemical Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

24. 72.01.03a · Additional evidence

    Misuse Risks

    Physical Harm and Injury Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

25. 72.01.04a · Additional evidence

    Misuse Risks

    Large-Scale Persuasion and Harmful Manipulation Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

26. 72.02.00a · Additional evidence

    Loss of Control Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

27. 72.02.02a · Additional evidence

    Loss of Control Risks

    Active loss of control

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

    Other Other

28. 72.02.02b · Additional evidence

    Loss of Control Risks

    Active loss of control

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

29. 72.03.00a · Additional evidence

    Accident Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

30. 72.03.00b · Additional evidence

    Accident Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

31. 72.04.00 · Risk Category

    Systemic Risks

    "Systemic risks emerge from widespread deployment of general-purpose AI beyond the risks directly posed by capabilities of individual models. These risks arise from structural mismatches between AI technology and existing social, economic, and institutional frameworks, creating vulnerabilities that transcend individual model-level interventions and require coordinated industry-wide and societal-level responses."

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

    Other Unintentional Post-deployment

32. 72.04.00a · Additional evidence

    Systemic Risks

    —

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

33. 72.06.00 · Risk Category

    Model Propensities

    -

    From Frontier AI Risk Management Framework (v1.0) (Tse2025)

    Not coded Not coded Not coded

34. 73.03.02a · Additional evidence

    Dual-Use Capabilities Enable Malicious Use and Misuse of LLMs

    Cybersecurity

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

35. 73.03.02b · Additional evidence

    Dual-Use Capabilities Enable Malicious Use and Misuse of LLMs

    Cybersecurity

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

36. 73.03.02c · Additional evidence

    Dual-Use Capabilities Enable Malicious Use and Misuse of LLMs

    Cybersecurity

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

37. 73.03.05a · Additional evidence

    Dual-Use Capabilities Enable Malicious Use and Misuse of LLMs

    Hazardous Biological and Chemical Technologies

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

38. 73.04.00 · Risk Category

    LLM-Systems Can Be Untrustworthy

    "A key desideratum for an LLM from a user’s perspective is ‘trustworthiness’, i.e. assurance of reliability and consistent performance, and absence of any accidental harm caused by the technology to the user.16 Providing assurance that an LLM-based system will not cause accidental harm remains a major open challenge. Harms may either occur directly due to the flawed nature of LLMs, e.g. an LLM generating toxic language or behaving inappropriately in some other ways, or may occur due to improper usage by a user, e.g. automation bias due to a user’s overreliance on LLM."

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

    AI Unintentional Post-deployment

39. 73.04.01a · Additional evidence

    LLM-Systems Can Be Untrustworthy

    Harms of Representation and Other Biases

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

40. 73.04.02a · Additional evidence

    LLM-Systems Can Be Untrustworthy

    Inconsistent Performance across and within Domains

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

41. 73.04.03a · Additional evidence

    LLM-Systems Can Be Untrustworthy

    Overreliance

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

42. 73.05.00 · Risk Category

    Socioeconomic Impacts of LLM May Be Highly Disruptive

    "The rapid evolution of LLMs brings significant socioeconomic opportunities and challenges, impacting the workforce, income inequality, education, and global economic development. Many of these challenges are systemic in nature, constituting what economists refer to as general equilibrium effects. These challenges do not arise directly from LLMs causing harm to users but rather from their indirect effects on the socioeconomic equilibrium."

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

    Other Unintentional Post-deployment

43. 73.05.00a · Additional evidence

    Socioeconomic Impacts of LLM May Be Highly Disruptive

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

44. 73.05.01a · Additional evidence

    Socioeconomic Impacts of LLM May Be Highly Disruptive

    Effects on the Workforce

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

45. 73.05.01b · Additional evidence

    Socioeconomic Impacts of LLM May Be Highly Disruptive

    Effects on the Workforce

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

46. 73.05.03a · Additional evidence

    Socioeconomic Impacts of LLM May Be Highly Disruptive

    Global Economic Development

    —

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

47. 73.07.03 · Risk Sub-Category

    Jailbreaks and Prompt Injections Threaten Security of LLMs

    Adversarial Optimization:

    "Jailbreak attacks can be discovered by performing manual or auto- mated adversarial optimization against a proxy objective that is noisily correlated with the success of a jailbreak. These are mostly gradient-based attacks (Zou et al., 2023b; Shin et al., 2020) as described in the previous two challenges, but gradient-free methods also exist (Prasad et al., 2022; Deng et al., 2022; Lapid et al., 2023)."

    From Foundational Challenges in Assuring Alignment and Safety of Large Language Models (Anwar2024)

    Not coded Not coded Not coded

48. 74.01.06a · Additional evidence

    Inherent Risk

    Hallucination

    —

    From A Survey on Responsible LLMs: Inherent Risk, Malicious Use, and Mitigation Strategy (Wang2025)

49. 74.01.07a · Additional evidence

    Inherent Risk

    Value-related risks in LLMs

    —

    From A Survey on Responsible LLMs: Inherent Risk, Malicious Use, and Mitigation Strategy (Wang2025)

50. 74.02.01a · Additional evidence

    Malicious Use

    Toxicity in LLM Malicious Use

    —

    From A Survey on Responsible LLMs: Inherent Risk, Malicious Use, and Mitigation Strategy (Wang2025)

Disclaimer: informational only, not legal advice. Verify every claim against the linked official sources and consult a qualified lawyer before acting.