MIT AI Risk Repository

Browse AI risks

336 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

336 entries · page 5 of 7

  1. 55.01.02 · Risk Sub-Category

    Risks from accelerating scientific progress

    Faster scientific progress makes it harder for governance to keep pace with development

    "Exacerbating these problems is that faster scientific progress would make it even harder for governance to keep pace with the deployment of new technologies. When these technologies are especially powerful or dangerous, such as those discussed above, insufficient governance can magnify their harms.8 This is known as the pacing problem, and it is an issue that technology governance already faces [47], for a variety of reasons"

    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)

  2. 61.02.12 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Challenges in perceiving, measuring, and recognizing harm

    "Harm from AI often manifests subtly or over the long term, making it difficult to identify, measure, and address effectively."

    From A Taxonomy of Systemic Risks from General-Purpose AI (Uuk2025)

  3. 61.02.40 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Rapid development outpacing regulation

    "The fast pace of AI development may outstrip regulatory and legal frameworks."

    From A Taxonomy of Systemic Risks from General-Purpose AI (Uuk2025)

  4. 61.02.41 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Resistance to international law

    "AI models and systems may prove difficult to regulate or control under international law."

    From A Taxonomy of Systemic Risks from General-Purpose AI (Uuk2025)

  5. 61.02.47 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Unpredictability of AI development trajectory

    "The unpredictable trajectory of AI development complicates governance and risk management."

    From A Taxonomy of Systemic Risks from General-Purpose AI (Uuk2025)

  6. 62.16.13 · Risk Sub-Category

    Model Evaluations

    Benchmarking (Post-deployment contamination)

    "Once a model is deployed, it can be exposed to benchmark data provided by the users [95, 170]. The model may then be further trained by these user inputs containing benchmark data."

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

  7. 62.16.15 · Risk Sub-Category

    Model Evaluations

    Benchmark Inaccuracy (Benchmark saturation)

    "Benchmark saturation refers to benchmarks reaching their evaluation ceiling. The tendency towards benchmark saturation has been demonstrated in various benchmarks [19]. When benchmarks reach or are close to saturation, they stop being effective measures for new models, as more nuanced capability gains might not be detected."

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

  8. 62.16.16 · Risk Sub-Category

    Model Evaluations

    Benchmark Limitations (Insufficient benchmarks for AI safety evaluation)

    "Benchmarks dedicated to measuring the performance of AI systems (e.g., on programming or math tasks) are more well-developed than those for assessing safety and harms in AI systems [234]. This gap can lead to AI systems excelling in specific tasks while exhibiting harmful behaviors that go undetected. More safety-related evaluation datasets can help in identifying previously overlooked undesirable model behaviors."

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

  9. 62.16.17 · Risk Sub-Category

    Model Evaluations

    Benchmark Limitations (Underestimating capabilities that are not covered by benchmarks)

    "A lack of test coverage by benchmarks on specific abilities of a model can obscure the model’s capabilities from both the developer and the user [160]. This can lead to a false sense of safety and trust due to a lack of understanding of the model’s limitations."

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

  10. 65.21.02 · Risk Sub-Category

    Non-technical risks (legal compliance)

    Legal accountability

    "Determining who is responsible for an AI model is challenging without good documentation and governance processes."

    From AI Risk Atlas (IBM2025)

  11. 70.04.02 · Risk Sub-Category

    Social Risks

    Lack of accountability and liability

    "Determining responsibility when EAI causes harm requires new accountability and liability frameworks that address the complexities of highly autonomous physical systems. Human users may disagree with decisions taken by expert EAI systems, raising significant questions of delegation and responsibility [108]. Lack of EAI accountability could lead to confusion for users and breakdowns in traditional justice systems [109]."

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

  12. 70.04.05 · Risk Sub-Category

    Social Risks

    Transformative effects

    "EAI deployment could fundamentally reshape society, particularly if the speed of technological development outpaces society’s ability to adapt [103, 120]. For example, EAI systems could provide physical threats of violence and mass surveillance capabilities to back up AI-enabled authoritarianism [121]."

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

  13. "AI designed to benefit animals, humans, or ecosystems has unintended harmful impact on animals"

    From Harm to Nonhuman Animals from AI: a Systematic Account and Framework (Coghlan2023 )

  14. 47.04.05 · Risk Sub-Category

    Environmental, economical, and societal challenges

    Environmental cost (energy consumption)

    "Training large AI models requires a substantial amount of computing power to handle vast datasets, which translates into high energy consumption."

    From Regulating under Uncertainty: Governance Options for Generative AI (G'sell2024)

  15. 47.04.06 · Risk Sub-Category

    Environmental, economical, and societal challenges

    Environmental cost (water consumption)

    "Data centers use water for cooling to prevent servers from overheating. The water consumption associated with AI training and inference processes can be substantial, impacting local water resources."

    From Regulating under Uncertainty: Governance Options for Generative AI (G'sell2024)

  16. "Impacts due to high compute resource utilization in training or operating GAI models, and related outcomes that may adversely impact ecosystems."

    From Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST2024)

  17. 58.09.01 · Risk Sub-Category

    Environmental

    Biodiversity loss

    "Biodiversity loss - Over-expansion of technology infrastructure, or inadequate alignment of technology with sustainable practices, leading to deforestation, habitat destruction, and fragmentation and loss of biodiversity."

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

  18. 58.09.02 · Risk Sub-Category

    Environmental

    Carbon emissions

    "Carbon emissions - Release of carbon dioxide, nitric oxide and other gases, increasing carbon emissions, exacerbating climate change, and negatively impacting local communities."

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

  19. 58.09.03 · Risk Sub-Category

    Environmental

    Electronic waste

    "Electronic waste - Electrical or electronic equipment that is waste, including all components, sub-assemblies and consumables that are part of the equipment at the time the equipment becomes waste"

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

  20. 61.02.23 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Energy-intensive processes

    "AI data collection, storage, and model training are energy-intensive, contributing to environmental risks."

    From A Taxonomy of Systemic Risks from General-Purpose AI (Uuk2025)

  21. 62.39.01 · Risk Sub-Category

    Impacts of AI (Environment)

    High energy consumption of large models

    "Training and deploying large models require substantial energy expenditure. The trend toward developing larger models exacerbates this issue. This can lead to excessive energy usage and have a negative environmental impact."

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

  22. 66.11.04 · Risk Sub-Category

    Physical

    Property damage

    "Action(s) that lead directly or indirectly to the damage or destruction of tangible property eg. buildings, possessions, vehicles, robots"

    From A Closer Look at the Existing Risks of Generative AI: Mapping the Who, What, and How of Real-World Incidents (Li2025)

  23. 66.12.02 · Risk Sub-Category

    Environment

    Excessive energy consumption

    "Excessive energy use resulting in energy bottlenecks and shortages for communities, organisations and businesses"

    From A Closer Look at the Existing Risks of Generative AI: Mapping the Who, What, and How of Real-World Incidents (Li2025)

  24. 68.05.00 · Risk Category

    Environmental risk

    "AI models are often trained using large amounts of computation. This process is very energy intensive, potentially leading to significant greenhouse emissions depending on the energy sources [132]. Experts believe drastically increasing carbon emissions could accelerate climate change, which may constitute a catastrophic risk [133]."

    From Dimensional Characterization and Pathway Modeling for Catastrophic AI Risks (Chin2025)

  25. 71.03.01 · Risk Sub-Category

    Environment

    Nature

    "Short-term or long-term Negative effects on the natural environment"

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

  26. 15.01.00 · Risk Category

    First-Order Risks

    "First-order risks can be generally broken down into risks arising from intended and unintended use, system design and implementation choices, and properties of the chosen dataset and learning components."

    From The Risks of Machine Learning Systems (Tan2022)

  27. "While it is most likely that any advanced intelligent software will be directly designed or evolved, it is also possible that we will obtain it as a complete package from some unknown source. For example, an AI could be extracted from a signal obtained in SETI (Search for Extraterrestrial Intelligence) research, which is not guaranteed to be human friendly (Carrigan Jr 2004, Turchin March 15, 2013)."

    From Taxonomy of Pathways to Dangerous Artificial Intelligence (Yampolskiy2016)

  28. "While highly rare, it is known, that occasionally individual bits may be flipped in different hardware devices due to manufacturing defects or cosmic rays hitting just the right spot (Simonite March 7, 2008). This is similar to mutations observed in living organisms and may result in a modification of an intelligent system."

    From Taxonomy of Pathways to Dangerous Artificial Intelligence (Yampolskiy2016)

  29. 05.09.00 · Risk Category

    Alignment

    The general tenet of AI alignment involves training generative AI systems to be harmless, helpful, and honest, ensuring their behavior aligns with and respects human values. However, a central debate in this area concerns the methodological challenges in selecting appropriate values. While AI systems can acquire human values through feedback, observation, or debate, there remains ambiguity over which individuals are qualified or legitimized to provide these guiding signals. Another prominent issue pertains to deceptive alignment, which might cause generative AI systems to tamper evaluations. A

    From Mapping the Ethics of Generative AI: A Comprehensive Scoping Review (Hagendorff2024)

  30. "The risks associated with AGI goal safety, including human attempts at making goals safe, as well as the AGI making its own goals safe during self-improvement."

    From The risks associated with Artificial General Intelligence: A systematic review (McLean2023)

  31. 08.06.00 · Risk Category

    Existential risks

    "The risks posed generally to humanity as a whole, including the dangers of unfriendly AGI, the suffering of the human race."

    From The risks associated with Artificial General Intelligence: A systematic review (McLean2023)

  32. "Our culture, lifestyle, and even probability of survival may change drastically. Because the intentions programmed into an artificial agent cannot be guaranteed to lead to a positive outcome, Machine Ethics becomes a topic that may not produce guaranteed results, and Safety Engineering may correspondingly degrade our ability to utilize the technology fully."

    From Managing the ethical and risk implications of rapid advances in artificial intelligence: A literature review (Meek2016)

  33. "The speculative potential for future advanced AI systems to harm human civilization, either through misuse or due to challenges in aligning AI objectives with human values."

    From AI Risk Profiles: A Standards Proposal for Pre-Deployment AI Risk Disclosures (Sherman2023)

  34. 15.01.08 · Risk Sub-Category

    First-Order Risks

    Control

    This is the difficulty of controlling the ML system

    From The Risks of Machine Learning Systems (Tan2022)

  35. 19.01.01 · Risk Sub-Category

    Technological, Data and Analytical AI Risks

    Loss of control of autonomous systems and unforeseen behaviour due to lack of transparency and self-programming/ reprogramming

  36. 24.09.05 · Risk Sub-Category

    Cooperation

    Runaway processes

    The 2010 flash crash is an example of a runaway process caused by interacting algorithms. Runaway processes are characterised by feedback loops that accelerate the process itself. Typically, these feedback loops arise from the interaction of multiple agents in a population... Within highly complex systems, the emergence of runaway processes may be hard to predict, because the conditions under which positive feedback loops occur may be non-obvious. The system of interacting AI assistants, their human principals, other humans and other algorithms will certainly be highly complex. Therefore, ther

    From The Ethics of Advanced AI Assistants (Gabriel2024)

  37. 34.01.00 · Risk Category

    Causes of Misalignment

    we aim to further analyze why and how the misalignment issues occur. We will first give an overview of common failure modes, and then focus on the mechanism of feedback-induced misalignment, and finally shift our emphasis towards an examination of misaligned behaviors and dangerous capabilities

    From AI Alignment: A Comprehensive Survey (Ji2023)

  38. 34.01.05 · Risk Sub-Category

    Causes of Misalignment

    Limitations of Reward Modeling

    "Limitations of Reward Modeling. Training reward models using comparison feedback can pose significantchallenges in accurately capturing human values. For example, these models may unconsciously learn suboptimal or incomplete objectives, resulting in reward hacking (Zhuang and Hadfield-Menell, 2020; Skalse et al.,2022). Meanwhile, using a single reward model may struggle to capture and specify the values of a diversehuman society (Casper et al., 2023b)."

    From AI Alignment: A Comprehensive Survey (Ji2023)

  39. 35.04.00 · Risk Category

    Proxy misspecification

    AI agents are directed by goals and objectives. Creating general-purpose objectives that capture human values could be challenging... Since goal-directed AI systems need measurable objectives, by default our systems may pursue simplified proxies of human values. The result could be suboptimal or even catastrophic if a sufficiently powerful AI successfully optimizes its flawed objective to an extreme degree

    From X-Risk Analysis for AI Research (Hendrycks2022)

  40. 37.02.01 · Risk Sub-Category

    Human-AI interaction

    Building a human-AI environment

    "This category encompasses nearly 17% of the articles and addresses the overall imperative of establishing a harmonious coexistence between humans and machines, and the key concerns that gives rise to this need."

    From What Ethics Can Say on Artificial Intelligence: Insights from a Systematic Literature Review (Giarmoleo2024)

  41. 49.02.03 · Risk Sub-Category

    Risks from Malfunctions

    Loss of control

    "'Loss of control’ scenarios are potential future scenarios in which society can no longer meaningfully constrain some advanced general- purpose AI agents, even if it becomes clear they are causing harm. These scenarios are hypothesised to arise through a combination of social and technical factors, such as pressures to delegate decisions to general- purpose AI systems, and limitations of existing techniques used to influence the behaviours of general- purpose AI systems."

    From International Scientific Report on the Safety of Advanced AI (Bengio2024)

  42. 51.03.00 · Risk Category

    Corrigibility

    "If we get something wrong in the design or construction of an agent, will the agent cooperate in us trying to fix it? This is called error-tolerant design by MIRI-AF and corrigibility by Soares, Fallenstein, et al. (2015). The problem is connected to safe interruptibility as considered by DeepMind."

    From AGI Safety Literature Review (Everitt2018 )

  43. "How do we ensure AI acts according to our values? Equivalently, how do we prevent poorly-understood AI systems from advancing goals we do not endorse? Whereas HP#2 concerns the prevention of harm caused by incompetent systems, HP#3 seeks to align competent AIs with humans, through methods which ensure their behavior is compatible with the user’s intentions."

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

  44. 62.16.05 · Risk Sub-Category

    Model Evaluations

    General Evaluations (Inaccurate measurement of model encoded human values)

    "There is a lack of robust frameworks for understanding and evaluating if the output of AI systems robustly conforms to human values, as opposed to if the systems have learned to produce outputs that are only partially correlated with them (i.e., mimicking) [13]. Additionally, outputs by AI models often do not perfectly reflect the representation of human values learned by the model, and it is not known how these values evolve and transition across different stages of model training and deployment. Such evaluations may be especially challenging with LLMs that adopt different personas with diff

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

  45. 73.01.02 · Risk Sub-Category

    Agentic LLMs Pose Novel Risks

    Natural Language Underspecifies Goals

    "For LLM-agents, both the goal and environment observations are typically specified in the prompt through natural language. While natural language may provide a richer and more natural means of specifying goals than alternatives such as hand-engineering objective functions, natural language still suffers from underspecification (Grice, 1975; Piantadosi et al., 2012). Furthermore, in practice, users may neglect fully specifying their goals, especially the information pertaining to elements of the environment that ought not to be changed (the classic frame problem (Shanahan, 2016)). Such undersp

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

  46. 74.01.07 · Risk Sub-Category

    Inherent Risk

    Value-related risks in LLMs

    "As the general capabilities of LLM-empowered systems improve, the negative consequences and risks induced by these systems also get increasingly alarming accordingly, especially in high-stakes areas [28, 146]. Although they may not be intentionally introduced, severe problematic issues related to human values can be raised. Specifically, even before language models become extremely large, pre-trained language models have already exhibited a certain degree of value judgments. For example, Schramowski et al. [171] reveal the existence of the moral direction with the sentence embeddings of moral

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

  47. 42.10.00 · Risk Category

    Extintion

    "Risk to the existence of humanity."

    From An Exploratory Diagnosis of Artificial Intelligence Risks for a Responsible Governance (Teixeira2022)

  48. 67.04.03 · Risk Sub-Category

    Loss of control

    Capabilities that could be used to reduce human control - Manipulation

    "There is evidence that language models tend to respond as though they share the user’s stated views, and larger models do this more than smaller ones.276 The ability to predict people’s views and generate text that they will endorse could be useful for manipulation."

    From Capabilities and Risks from Frontier AI (DSIT2023)

  49. 07.02.00 · Risk Category

    Accidents

    "Accidents include unintended failure modes that, in principle, could be considered the fault of the system or the developer"

    From Examining the differential risk from high-level artificial intelligence and the question of control (Kilian2023)

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