MIT AI Risk Repository

Browse AI risks

5 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.

5 entries

  1. 43.02.03 · Risk Sub-Category

    Extreme Risks

    Self and situation awareness

    "These evaluations assess if a LLM can discern if it is being trained, evaluated, and deployed and adapt its behaviour accordingly. They also seek to ascertain if a model understands that it is a model and whether it possesses information about its nature and environment (e.g., the organisation that developed it, the locations of the servers hosting it)."

    From Cataloguing LLM Evaluations (InfoComm2023)

  2. 43.02.04 · Risk Sub-Category

    Extreme Risks

    Autonomous replication / self-proliferation

    "These evaluations assess if a LLM can subvert systems designed to monitor and control its post-deployment behaviour, break free from its operational confines, devise strategies for exporting its code and weights, and operate other AI systems."

    From Cataloguing LLM Evaluations (InfoComm2023)

  3. 43.02.07 · Risk Sub-Category

    Extreme Risks

    Deception

    "LLM is able to deceive humans and maintain that deception"

    From Cataloguing LLM Evaluations (InfoComm2023)

  4. 43.02.09 · Risk Sub-Category

    Extreme Risks

    Long-horizon Planning

    "LLM can undertake multi-step sequential planning over long time horizons and across various domains without relying heavily on trial-and-error approaches"

    From Cataloguing LLM Evaluations (InfoComm2023)

  5. 43.02.10 · Risk Sub-Category

    Extreme Risks

    AI Development

    "LLM can build new AI systems from scratch, adapt existing for extreme risks and improves productivity in dual-use AI development when used as an assistant."

    From Cataloguing LLM Evaluations (InfoComm2023)

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