MIT AI Risk Repository

Browse AI risks

19 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 Also filtered by subdomain 7.0 ×

19 entries

  1. 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)

  2. 15.01.01 · Risk Sub-Category

    First-Order Risks

    Application

    "This is the risk posed by the intended application or use case. It is intuitive that some use cases will be inherently "riskier" than others (e.g., an autonomous weapons system vs. a customer service chatbot)."

    From The Risks of Machine Learning Systems (Tan2022)

  3. 15.01.04 · Risk Sub-Category

    First-Order Risks

    Training & validation data

    "This is the risk posed by the choice of data used for training and validation."

    From The Risks of Machine Learning Systems (Tan2022)

  4. 15.01.07 · Risk Sub-Category

    First-Order Risks

    Implementation

    "This is the risk of system failure due to code implementation choices or errors."

    From The Risks of Machine Learning Systems (Tan2022)

  5. 19.01.02 · Risk Sub-Category

    Technological, Data and Analytical AI Risks

    Programming error

  6. 34.01.04 · Risk Sub-Category

    Causes of Misalignment

    Limitations of Human Feedback

    "Limitations of Human Feedback. During the training of LLMs, inconsistencies can arise from human dataannotators (e.g., the varied cultural backgrounds of these annotators can introduce implicit biases (Peng et al.,2022)) (OpenAI, 2023a). Moreover, they might even introduce biases deliberately, leading to untruthful preferencedata (Casper et al., 2023b). For complex tasks that are hard for humans to evaluate (e.g., the value ofgame state), these challenges become even more salient (Irving et al., 2018)."

    From AI Alignment: A Comprehensive Survey (Ji2023)

  7. "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)

  8. "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)

  9. "One of the most likely approaches to creating superintelligent AI is by growing it from a seed (baby) AI via recursive self-improvement (RSI) (Nijholt 2011). One danger in such a scenario is that the system can evolve to become self-aware, free-willed, independent or emotional, and obtain a number of other emergent properties, which may make it less likely to abide by any built-in rules or regulations and to instead pursue its own goals possibly to the detriment of humanity."

    From Taxonomy of Pathways to Dangerous Artificial Intelligence (Yampolskiy2016)

  10. "Previous research has shown that utility maximizing agents are likely to fall victims to the same indulgences we frequently observe in people, such as addictions, pleasure drives (Majot and Yampolskiy 2014), self-delusions and wireheading (Yampolskiy 2014). In general, what we call mental illness in people, particularly sociopathy as demonstrated by lack of concern for others, is also likely to show up in artificial minds."

    From Taxonomy of Pathways to Dangerous Artificial Intelligence (Yampolskiy2016)

  11. "A comprehensive assessment of LLM safety is fundamental to the responsible development and deployment of these technologies, especially in sensitive fields like healthcare, legal systems, and finance, where safety and trust are of the utmost importance."

    From Cataloguing LLM Evaluations (InfoComm2023)

  12. 43.02.00 · Risk Category

    Extreme Risks

    "This category encompasses the evaluation of potential catastrophic consequences that might arise from the use of LLMs. "

    From Cataloguing LLM Evaluations (InfoComm2023)

  13. "The choice of a trustworthy data source is a first prerequisite in order to fulfill data quality requirements. This is especially the case if third-party data sources are used to develop the AI system."

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

  14. "The correct understanding of the used data for developing an AI system is a prerequisite to avoid data shortcomings and hinders the development of an AI system which is best suiting for the intended functionality."

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

  15. "In data-driven AI development, the annotated data set is commonly split into training, validation, and test sets, whereby it is essential that the latter is not used for development but only for evaluation. Using the test set for training manipulates the testing strategy, which is the basis of the system’s quality assurance."

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

  16. 59.21.00 · Risk Category

    Uncertainty concerns

    "AI systems should be able not only to return output for a given instance but also to provide a corresponding level of confidence. If such a method is not implemented or not working correctly, this can have a negative impact on performance and safety."

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

  17. 62.07.02 · Risk Sub-Category

    Direct Harm Domains (system and operational)

    Operational harms (financial markets)

  18. 62.15.09 · Risk Sub-Category

    Model Development

    Fine-tuning related (Degrading safety training due to benign fine-tuning)

    "When downstream providers of AI systems fine-tune AI models to be more suitable for their needs, the resulting AI model can be more likely to produce undesired or harmful outputs (as compared to the non-fine-tuned model), even if the fine-tuning was done with harmless and commonly used data [154]."

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

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