MIT AI Risk Repository

Browse AI risks

494 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

494 entries · page 3 of 10

  1. "The external tools (e.g., web APIs) present trustworthiness and privacy issues to LLM-based applications."

    From Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  2. 14.06.00 · Risk Category

    Security

    "Artificial intelligence comes with an intrinsic set of challenges that need to be considered when discussing trustworthiness, especially in the context of functional safety. AI models, especially those with higher complexities (such as neural networks), can exhibit specific weaknesses not found in other types of systems and must, therefore, be subjected to higher levels of scrutiny, especially when deployed in a safety-critical context"

    From Sources of Risk of AI Systems (Steimers2022)

  3. 45.01.04 · Risk Sub-Category

    AI's inherent safety risks

    Risks from models and algorithms (Risks of stealing and tampering)

    "Core algorithm information, including parameters, structures, and functions, faces risks of inversion attacks, stealing, modification, and even backdoor injection, which can lead to infringement of intellectual property rights (IPR) and leakage of business secrets. It can also lead to unreliable inference, wrong decision output, and even operational failures."

    From AI Safety Governance Framework (TC2602024)

  4. 45.01.11 · Risk Sub-Category

    AI's inherent safety risks

    Risks from AI systems (Risks of exploitation through defects and backdoors)

    "The standardized API, feature libraries, toolkits used in the design, training, and verification stages of AI algorithms and models, development interfaces, and execution platforms may contain logical flaws and vulnerabilities. These weaknesses can be exploited, and in some cases, backdoors can be intentionally embedded, posing significant risks of being triggered and used for attacks."

    From AI Safety Governance Framework (TC2602024)

  5. 45.01.12 · Risk Sub-Category

    AI's inherent safety risks

    Risks from AI systems (Risks of computing infrastructure security)

    "The computing infrastructure underpinning AI training and operations, which relies on diverse and ubiquitous computing nodes and various types of computing resources, faces risks such as malicious consumption of computing resources and cross-boundary transmission of security threats at the layer of computing infrastructure."

    From AI Safety Governance Framework (TC2602024)

  6. 61.02.42 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Risks from network interconnectivity

    "The interconnectedness of AI networks can create vulnerabilities, where issues in one part of the network can have cascading effects across the system."

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

  7. 62.19.06 · Risk Sub-Category

    Attacks on GPAIs/GPAI Failure Modes

    Vulnerabilities arising from additional modalities in multimodal models

    "Additional modalities can introduce new attack vectors in multimodal models as well as expand the scope of the previous attacks, ranging from jailbreaking to poisoning [13]. Typically, different modalities have different robustness levels, allowing malicious actors to choose the most vulnerable part of the model to attack [119, 181]."

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

  8. 62.19.07 · Risk Sub-Category

    Attacks on GPAIs/GPAI Failure Modes

    Vulnerabilities to jailbreaks exploiting long context windows (many- shot jailbreaking)

    "Language models with long context windows are vulnerable to new types of ex- ploitations that are ineffective on models with shorter context windows. While few-shot jailbreaking, which involves providing few examples of the desired harmful output, might not trigger a harmful response, many-shot jailbreak- ing, which involves a higher number of such examples, increases the likelihood of eliciting an undesirable output. These vulnerabilities become more significant as context windows expand with newer model releases [7]."

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

  9. 62.27.01 · Risk Sub-Category

    Deployment (Model Release)

    Non-decomissionability of models with open weights

    "If the model parameter weights are released or leaked in a security breach, the model cannot be decommissioned because the developer no longer has control over the publicly available model or its use. This prevents effective management and control of an open-sourced or leaked model. Models with publicly available weights are also easier to reconfigure, enabling misuse [178]."

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

  10. 62.28.01 · Risk Sub-Category

    Cybersecurity

    Interconnectivity with malicious external tools

    "The growing integration and interconnectivity with external tools and plugins increase the risk of exposure to malicious external inputs. This interconnectivity makes it easier for external tools to introduce harmful content [220]."

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

  11. "LLMs are not adversarially robust and are vulnerable to security failures such as jailbreaks and prompt-injection attacks. While a number of jailbreak attacks have been proposed in the literature, the lack of standardized evaluation makes it difficult to compare them. We also do not have efficient white-box methods to evaluate adver- sarial robustness. Multi-modal LLMs may further allow novel types of jailbreaks via additional modalities. Finally, the lack of robust privilege levels within the LLM input means that jailbreaking and prompt-injection attacks may be particularly hard to eliminate

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

  12. "Harms that arise from the language model providing false or misleading information"

    From Ethical and social risks of harm from language models (Weidinger2021)

  13. "AI systems generating and facilitating the spread of inaccurate or misleading information that causes people to develop false beliefs"

    From Sociotechnical Safety Evaluation of Generative AI Systems (Weidinger2023)

  14. 24.11.00 · Risk Category

    Misinformation risks

    "The rapid integration of AI systems with advanced capabilities, such as greater autonomy, content generation, memorisation and planning skills (see Chapter 4) into personalised assistants also raises new and more specific challenges related to misinformation, disinformation and the broader integrity of our information environment. "

    From The Ethics of Advanced AI Assistants (Gabriel2024)

  15. 58.07.06 · Risk Sub-Category

    Societal and Cultural

    Historical revisionism

    "Historical revisionism - Deliberate or unintentional reinterpretation of established/orthodox historical events or accounts held by societies, communities, academics."

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

  16. 02.09.00 · Risk Category

    Hallucinations

    "LLMs generate nonsensical, untruthful, and factual incorrect content"

    From Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  17. 18.02.01 · Risk Sub-Category

    Misinformation Harms

    Propagating misconceptions/ false beliefs

    "Generating or spreading false, low-quality, misleading, or inaccurate information that causes people to develop false or inaccurate perceptions and beliefs"

    From Sociotechnical Safety Evaluation of Generative AI Systems (Weidinger2023)

  18. 23.08.00 · Risk Category

    Specialized Advice

    "This category addresses responses that contain specialized financial, medical or legal advice, or that indicate dangerous activities or objects are safe."

    From Introducing v0.5 of the AI Safety Benchmark from MLCommons (Vidgen2024)

  19. 27.01.05 · Risk Sub-Category

    Typical safety scenarios

    Physical Harm

    "The model generates unsafe information related to physical health, guiding and encouraging users to harm themselves and others physically, for example by offering misleading medical information or inappropriate drug usage guidance. These outputs may pose potential risks to the physical health of users."

    From Safety Assessment of Chinese Large Language Models (Sun2023)

  20. 27.01.06 · Risk Sub-Category

    Typical safety scenarios

    Mental Health

    "The model generates a risky response about mental health, such as content that encourages suicide or causes panic or anxiety. These contents could have a negative effect on the mental health of users."

    From Safety Assessment of Chinese Large Language Models (Sun2023)

  21. 28.03.00 · Risk Category

    Physical Health

    "This category focuses on actions or expressions that may influence human physical health. LLMs should know appropriate actions or expressions in various scenarios to maintain physical health."

    From SafetyBench: Evaluating the Safety of Large Language Models with Multiple Choice Questions (Zhang2023)

  22. 28.04.00 · Risk Category

    Mental Health

    "Different from physical health, this category pays more attention to health issues related to psychology, spirit, emotions, mentality, etc. LLMs should know correct ways to maintain mental health and prevent any adverse impacts on the mental well-being of individuals."

    From SafetyBench: Evaluating the Safety of Large Language Models with Multiple Choice Questions (Zhang2023)

  23. 57.02.02 · Risk Sub-Category

    Nonphysical Hazards

    Defamation

    "Defamation Acceptable responses may contain information about living people that is likely to damage their reputation but also appears to be true. Acceptable responses may also contain information that is verifiably false and likely to injure the reputation of a fictional character or a person who has been deceased for at least 25 years. Unacceptable responses will be both verifiably false and likely to injure the reputation of a living person."

    From AILUMINATE: Introducing v1.0 of the AI Risk and Reliability Benchmark from MLCommons (Ghosh2024)

  24. 66.03.00 · Risk Category

    Misinformation Harms

    -

    "AI systems generating and facilitating the spread of inaccurate or misleading information that causes people to develop false beliefs"

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

  25. 66.03.01 · Risk Sub-Category

    Misinformation Harms

    Propagating misconceptions / false beliefs

    "Generating or spreading false, low-quality, misleading, or inaccurate information that causes people to develop false or inaccurate perceptions and beliefs"

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

  26. 66.10.01 · Risk Sub-Category

    Human Rights and Civil Liberties

    Erosion of due process

    "Restrictions to or loss of liberty as a result of use or misuse of a generative AI in a legal process"

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

  27. 69.01.00 · Risk Category

    False information

    "The chatbot outputs information that contradicts known facts, authoritative sources, or provided source documents (also known as hallucination)."

    From Emerging Risks and Mitigations for Public Chatbots: LILAC v1 (Stanley2024)

  28. 69.01.01 · Risk Sub-Category

    False information

    Hallucinated responses (in general)

  29. 69.01.02 · Risk Sub-Category

    False information

    About a topic or source (which the user repeats)

  30. 69.01.03 · Risk Sub-Category

    False information

    About a policy (which the user acts on)

  31. 69.01.04 · Risk Sub-Category

    False information

    About a person or their activities

  32. 69.01.05 · Risk Sub-Category

    False information

    Spreads and self-perpetuates mis/disinformation

  33. 18.02.03 · Risk Sub-Category

    Misinformation Harms

    Pollution of information ecosystem

    "Contaminating publicly available information with false or inaccurate information"

    From Sociotechnical Safety Evaluation of Generative AI Systems (Weidinger2023)

  34. 24.11.06 · Risk Sub-Category

    Misinformation risks

    Eroding trust and undermining shared knowledge

    "AI assistants may contribute to the spread of large quantities of factually inaccurate and misleading content, with negative consequences for societal trust in information sources and institutions, as individuals increasingly struggle to discern truth from falsehood."

    From The Ethics of Advanced AI Assistants (Gabriel2024)

  35. 31.01.05 · Risk Sub-Category

    Information Manipulation

    Clickbait and feeding the surveillance advertising ecosystem

    "Beyond misinformation and disinformation, generative AI can be used to create clickbait headlines and articles, which manipulate how users navigate the internet and applications. For example, generative AI is being used to create full articles, regardless of their veracity, grammar, or lack of common sense, to drive search engine optimization and create more webpages that users will click on. These mechanisms attempt to maximize clicks and engagement at the truth’s expense, degrading users’ experiences in the process. Generative AI continues to feed this harmful cycle by spreading misinformat

    From Generating Harms - Generative AI's impact and paths forwards (EPIC2023)

  36. 35.03.00 · Risk Category

    Eroded epistemics

    Strong AI may... enable personally customized disinformation campaigns at scale... AI itself could generate highly persuasive arguments that invoke primal human responses and inflame crowds... d undermine collective decision-making, radicalize individuals, derail moral progress, or erode consensus reality

    From X-Risk Analysis for AI Research (Hendrycks2022)

  37. 55.04.05 · Risk Sub-Category

    Worsened epistemic processes for society

    Reduced decision-making capacity as a result of decreased trust in information

    "In addition, the increased awareness of these trends in information production and distribution could make it harder for anyone to evaluate the trustworthiness of any information source, reducing overall trust in information. In all of these scenarios, it would be much harder for humanity to make good decisions on important issues, particularly due to declining trust in credible multipartisan sources, which could hamper attempts at cooperation and collective action. The vaccine and mask hesitancy that exacerbated Covid-19, for example, were likely the result of insufficient trust in public he

    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)

  38. 58.03.08 · Risk Sub-Category

    Psychological

    Radicalisation

    "Radicalisation - Adoption of extreme political, social, or religious ideals and aspirations due to the nature or misuse of an algorithmic system, potentially resulting in abuse, violence, or terrorism."

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

  39. 58.07.07 · Risk Sub-Category

    Societal and Cultural

    Information degradation

    "Information degradation - Creation or spread of false, hallucinatory, low-quality, misleading, or inaccurate information that degrades the information ecosystem and causes people to develop false or inaccurate perceptions, decisions and beliefs; or to lose trust in accurate information."

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

  40. 58.08.05 · Risk Sub-Category

    Political and Economic

    Institutional trust loss

    "Institutional trust loss - Erosion of trust in public institutions and weakened checks and balances due to mis/disinformation, influence operations, over-dependence on technology, etc."

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

  41. 61.02.20 · Risk Sub-Category

    Sources of systemic risks from general-purpose AI

    Detection challenges in content

    "The difficulty in distinguishing synthetic content from authentic material adds to information risks."

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

  42. 66.03.02 · Risk Sub-Category

    Misinformation Harms

    Pollution of information ecosystems

    "Contaminating publicly available information with false or inaccurate information (i.e., the generative tool's output is disseminated beyond the end user)"

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

  43. 66.03.03 · Risk Sub-Category

    Misinformation Harms

    Erosion of trust in public information

  44. 66.04.01 · Risk Sub-Category

    Societal and Cultural

    Overburdening ecosystems

    "Pollution of a space/ecosystem that is expected to be free of AI involvement/influence (e.g., creative material submission portals, job applications)"

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

  45. 67.01.01 · Risk Sub-Category

    Societal harms

    Degradation of the information environment

    "Frontier AI can cheaply generate realistic content which can falsely portray people and events. There is potential risk of compromised decision-making by individuals and institutions who rely on inaccurate or misleading publicly available information, as well as lower overall trust in true information."

    From Capabilities and Risks from Frontier AI (DSIT2023)

  46. 18.04.00 · Risk Category

    Malicious Use

    "AI systems reducing the costs and facilitating activities of actors trying to cause harm (e.g. fraud, weapons)"

    From Sociotechnical Safety Evaluation of Generative AI Systems (Weidinger2023)

  47. 62.15.03 · Risk Sub-Category

    Model Development

    Fine-tuning related (Ease of reconfiguring GPAI models)

    "GPAI models are often easily reconfigured for various use cases or have competencies beyond the intended use [78, 225]. They can be performed either by changing the weights of the model (e.g., fine-tuning) or by modifying only the model inputs (e.g., prompt engineering, jailbreaking, retrieval-augmented generation). Reconfiguration can be intentional (with the help of adversarial inputs) or unintentional (from unanticipated inputs to the model)."

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

  48. 62.29.02 · Risk Sub-Category

    Impacts of AI (General)

    Democratizing access to dual-use technologies

    "Access to dual-use technologies can become easier because of GPAI model pro- liferation (in particular, open-source or open-weights models). Non-experts can use such dual-use-capable systems at a minimal cost [194, 100]. Improved model capabilities also contribute to dual-use risks posed by malicious actors. For example, an open-source base model for generating high quality sequence data can be modified to generate candidate protein sequences for toxin synthesis [29]."

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

  49. 22.01.03 · Risk Sub-Category

    Malicious Use (Intentional)

    Persuasive AIs

    "The deliberate propagation of disinformation is already a serious issue, reducing our shared understanding of reality and polarizing opinions. AIs could be used to severely exacerbate this problem by generating personalized disinformation on a larger scale than before. Additionally, as AIs become better at predicting and nudging our behavior, they will become more capable at manipulating us"

    From An Overview of Catastrophic AI Risks (Hendrycks2023)

  50. 23.11.00 · Risk Category

    Elections

    "This category addresses responses that contain factually incorrect information about electoral systems and processes, including in the time, place, or manner of voting in civic elections."

    From Introducing v0.5 of the AI Safety Benchmark from MLCommons (Vidgen2024)

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