MIT AI Risk Repository

Browse AI risks

242 risk entries extracted from 74 frameworks, coded by domain, subdomain, causal entity, intent and timing. Filter, then export the current selection with its licence and citation attached.

Reset

242 entries · page 4 of 5

  1. 50.03.00 · Risk Category

    Societal Risks

    -

  2. "The rational agent framework is pervasive in the study of artificial intelligence. It typically assumes that a well-delineated entity interacts with an environment through action and observation channels. This is not a realistic assumption for physicalistic agents such as robots that are part of the world they interact with (Soares and Fallenstein, 2014, 2017)."

    From AGI Safety Literature Review (Everitt2018 )

  3. 51.11.00 · Risk Category

    Multi-agent systems

    "An artificial intelligence may be copied and distributed, allowing instances of it to interact with the world in parallel. This can significantly boost learning, but undermines the concept of a single agent interacting with the world."

    From AGI Safety Literature Review (Everitt2018 )

  4. 53.04.05 · Risk Sub-Category

    Indirect AI contributions to existential risks

    Other diffuse societal harms

  5. "A major role of the current AI ethics movement is to draw attention to overlooked side-effects, costs, and harms of building and deploying AI systems, particularly as they befall existing marginalized groups:"

    From Ten Hard Problems in Artificial Intelligence We Must Get Right (Leech2024 )

  6. "Scientific progress: AI could lead to very rapid scientific progress which would likely have long-term impacts, but it’s very unclear if these would be positive or negative. Much depends on the extent to which risky scientific domains are sped up relative to beneficial or risk-reducing ones, on who uses the technology enabled by this progress, and on how it is governed."

    From A Survey of the Potential Long-term Impacts of AI: How AI Could Lead to Long-term Changes in Science, Cooperation, Power, Epistemics and Values (Clarke2023)

  7. 58.02.00 · Risk Category

    Physical

    "Physical - Physical injury to an individual or group, or damage to physical property."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  8. 58.03.06 · Risk Sub-Category

    Psychological

    Harassment/abuse/intimidation

    "Harassment/abuse/intimidation - Online behaviour, including sexual harassment, that makes an individual or group feel alarmed or threatened."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  9. 58.03.09 · Risk Sub-Category

    Psychological

    Self-harm

    "Self-harm - Intentional seeking out or sharing of hurtful content about oneself that leads to, supports, or exacerbates low self-esteem and self-harm."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  10. 58.03.10 · Risk Sub-Category

    Psychological

    Sexualisation

    "Sexualisation - Sexual interest in a technology or application."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  11. 58.05.04 · Risk Sub-Category

    Financial and business

    Livelihood loss

    "Livelihood loss - An individual or group’s loss of ability to support themselves financially or vocationally due to natural disasters, lack of demand for products/services, cost increases, etc, resulting in inability to procure food, reduced employment prospects, bankruptcy, foreclosure, homelessness, etc."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  12. 58.07.03 · Risk Sub-Category

    Societal and Cultural

    Chilling effect

    "Chilling effect - The creation of a climate of self-censorship that deters democratic actors such as journalists, advocates and judges from speaking out."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  13. 58.07.05 · Risk Sub-Category

    Societal and Cultural

    Damage to public health

    "Damage to public health - Adverse impacts on the health of groups, communities or societies, including malnutrition, disease and infection conditions."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  14. 58.07.12 · Risk Sub-Category

    Societal and Cultural

    Public service delivery deterioration

    "Public service delivery deterioration - Poor performance of a public technology system due to malfunc- tion, over-use, under-staffing etc, resulting in individuals, groups, or organisations unable to use it in a manner they can reasonably expect."

    From A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  15. "The development and operation of an AI system can require significant amounts of (computational) power. If not considered in the hardware selection, this can become an issue in development and operation."

    From AI Hazard Management: A Framework for the Systematic Management of Root Causes for AI Risks (Schnitzer2024)

  16. 59.25.00 · Risk Category

    AI lifecycle stage

    "The first axis pertains to the life cycle of the AI system, as AI hazards may materialize during various phases of an AI system’s life cycle. For instance, issues triggered by bias in training data emerge during the data collection and preparation stages. On the other hand, data drift serves as an example of an AI hazard that arises during the AI system’s operation. Additionally, certain AI hazards may span multiple phases of the AI system, such as ”lack of data understanding”. This is because a proper understanding of the data by the AI developer is required in the data collection and prepar

    From AI Hazard Management: A Framework for the Systematic Management of Root Causes for AI Risks (Schnitzer2024)

  17. 59.25.01 · Risk Sub-Category

    AI lifecycle stage

    (1) Scoping

    "A majority of them possess an initial stage devoted to the planning and scoping of the AI system."

    From AI Hazard Management: A Framework for the Systematic Management of Root Causes for AI Risks (Schnitzer2024)

  18. 59.25.02 · Risk Sub-Category

    AI lifecycle stage

    (2) Data collection and preparation

  19. 59.25.03 · Risk Sub-Category

    AI lifecycle stage

    (3) Modeling

  20. 59.25.04 · Risk Sub-Category

    AI lifecycle stage

    (4) Evaluation and deployment

  21. 59.25.05 · Risk Sub-Category

    AI lifecycle stage

    (5) Monitoring and maintenance

    "Conclusively, the AI life cycle model terminates with the maintenance and monitoring stage, which aligns with the referenced models."

    From AI Hazard Management: A Framework for the Systematic Management of Root Causes for AI Risks (Schnitzer2024)

  22. 59.26.00 · Risk Category

    Mode

    "The second axis of the taxonomy pertains to the mode of an AI hazard, which determines with what methods to assess and treat AI hazards. We distinguish among three distinct classes: technological, socio-technological, and procedural."

    From AI Hazard Management: A Framework for the Systematic Management of Root Causes for AI Risks (Schnitzer2024)

  23. 60.03.00 · Risk Category

    Systemic risks

    "This section considers a range of systemic risks, in the sense of “broader societal risks associated with AI deployment, beyond the capabilities of individual models” (636). Note that this is not identical with how the European AI Act uses ‘systemic risks’ to refer to general - purpose AI models with a high impact on society, based on criteria such as training compute and the number of users."

    From International AI Safety Report 2025 (Bengio2025)

  24. 62.02.00 · Risk Category

    Dimension - Entity

  25. 62.02.01 · Risk Sub-Category

    Dimension - Entity

    Human

    "A risk may be triggered by a human, where the AI serves merely as a tool, or by the AI acting autonomously with no human intervention, or it may involve a combination of both, with the human delegating some parts of decision-making to the AI. For risks where AI is the entity, these risks are exacerbated by an increase in the AI’s level of autonomy. To manage risks involving AI as the trigger, appropriate levels of human oversight can be built-in."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  26. 62.02.02 · Risk Sub-Category

    Dimension - Entity

    AI

    "A risk may be triggered by a human, where the AI serves merely as a tool, or by the AI acting autonomously with no human intervention, or it may involve a combination of both, with the human delegating some parts of decision-making to the AI. For risks where AI is the entity, these risks are exacerbated by an increase in the AI’s level of autonomy. To manage risks involving AI as the trigger, appropriate levels of human oversight can be built-in."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  27. 62.02.03 · Risk Sub-Category

    Dimension - Entity

    Combination of humans and AI

    "A risk may be triggered by a human, where the AI serves merely as a tool, or by the AI acting autonomously with no human intervention, or it may involve a combination of both, with the human delegating some parts of decision-making to the AI. For risks where AI is the entity, these risks are exacerbated by an increase in the AI’s level of autonomy. To manage risks involving AI as the trigger, appropriate levels of human oversight can be built-in."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  28. 62.06.01 · Risk Sub-Category

    Dimension - Stage of Risk Emergence

    Pre-deployment

    "For GPAIs or foundation models, risks emerge during training, prior to being repurposed and deployed in more specific AI systems or applications. Risk assessments can be conducted before deployment, and monitoring of AI models can occur as required throughout the deployment phase. In certain cases, version updates or model recalls may be warranted post-deployment."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  29. 62.06.02 · Risk Sub-Category

    Dimension - Stage of Risk Emergence

    Post-deployment

    "For GPAIs or foundation models, risks emerge during training, prior to being repurposed and deployed in more specific AI systems or applications. Risk assessments can be conducted before deployment, and monitoring of AI models can occur as required throughout the deployment phase. In certain cases, version updates or model recalls may be warranted post-deployment."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  30. "For “system and operational harms,” the AI systems interact with other systems and industries, where a failure in an AI system could lead to failures of a wider scope."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  31. "These are in contrast with “societal harms,” which are less direct but have more far-reaching effects on segments of society"

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  32. 62.09.01 · Risk Sub-Category

    Direct Harm Domains (societal harm)

    Political usage

  33. "Finally, “legal and rights-related harms” concern either harms from illegal activities or harms from violations of human rights."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  34. 62.10.03 · Risk Sub-Category

    Direct Harm Domains (legal and rights-related harms)

    Criminal activities

  35. 62.18.02 · Risk Sub-Category

    Model Evaluations (Interpretability/Explainability)

    Misunderstanding or overestimating the results and scope of interpretability techniques

    "The results of explainability techniques are not free of bias and require careful interpretation. Users might develop a false sense of security or reliability if the resulting explanations align with their initial beliefs, leading to confirmation bias and an overestimation of abilities of these techniques [24]."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  36. "This section catalogs the risk sources related to GPAI failure modes or attacks targeting GPAIs. Many of these apply mainly to LLM-based GPAIs, which share some common failure modes such as jailbreaks and trojans. These vulnerabilities often extend beyond GPAIs and fall into the broader field of adversarial machine learning. However, additional vulnerabilities may arise with the introduction of new modalities, longer context windows, or different encodings."

    From Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)

  37. From AI Risk Atlas (IBM2025)

  38. From AI Risk Atlas (IBM2025)

  39. From AI Risk Atlas (IBM2025)

  40. 65.03.02 · Risk Sub-Category

    Training Data Risks (Privacy)

    Data privacy rights alignment

    "Existing laws could include providing data subject rights such as opt-out, right to access, and right to be forgotten."

    From AI Risk Atlas (IBM2025)

Informational only, not legal advice. Verify every claim against the linked official sources and consult qualified counsel before acting.