Saihub.info
Harms
Updated August 20, 2024
Harms that can be caused by artificial intelligence are a subject of intense debate. There is little agreement on what the potential harms are, or what harms are of greatest concern. The core mission of Saihub.info is to take a disciplined approach to identifying and analyzing all actual or potential harms from AI that are credible and significant. We list the types of harm that we have identified in our harms register, which is explained further below.
Harms-Specific, Risk-Based Approach
The debate about AI risk and AI safety has become confused by the tendency to treat disparate AI harms as a single problem, and by public disagreement over which harms deserve most attention (particularly between those focused on harms emerging now and those focused on long-term existential risk).
We believe that it is important to assess and respond to different harms differently, and often individually. We have begun to assess harms in this way, through:
There are various emerging standards and policies that inform risk management for AI:
ISO 23894 and the Artificial Intelligence Risk Management Framework (issued by the US National Institute of Standards and Technology (NIST)) set out AI-specific risk management approaches. As a supplement to its framework, NIST in April 2024 proposed the Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile.
Others are also urging a disciplined, risk-focused approach to AI harms:
A June 2024 letter by former and current employees of OpenAI and Google DeepMind (and endorsed by Yoshua Bengio, Geoff Hinton and Stuart Russell) advocates adoption by companies of an employee "right to warn" about the dangers of advanced AI technologies.
Our application of this harms/risk-specific approach is preliminary and incomplete, and we expect some of our harm/risk assessments to be controversial. Please get in touch if you have comments or suggested additions or changes to the harms register.
Closed vs Open Models
In discussions of AI harms / risks, an important cross-cutting issue (affecting multiple harms) is whether open-source AI models are beneficial (including because they promote innovation, and help avoid market capture by the largest companies) or dangerous (because they put the most powerful AI tools in the hands of many). An excellent paper on this issue was published by Stanford HAI in December 2023.
Other Resources
A few other resources on AI harms include:
the OECD AI Policy Observatory, which organizes by policy area the "latest AI policy research taking place in different policy communities across the OECD and beyond"
AI Safety Summits
series of reports on capabilities and risks of "frontier AI", published by the UK government ahead of the first AI Safety Summit in November 2023
International Scientific Report on the Safety of Advanced AI, published in connection with the Seoul AI Summit in May 2024
repositories of AI "incidents", which include taxonomies of incident types / harms
the GPT-4 System Card, which covers (1) "safety challenges presented by [GPT-4]", (2) "a high-level overview of the safety processes OpenAI adopted to prepare GPT-4 for deployment" and (3) a "demonstrat[ion] that while ... mitigations and processes alter GPT-4’s behavior and prevent certain kinds of misuses, they are limited and remain brittle in some cases"
Google DeepMind
Massachusetts Institute of Technology, AI Risk Repository (released August 2024)
Partnership on AI, Risk Mitigation Strategies for the Open Foundation Model Value Chain (July 2024)
Yoshua Bengio, Geoff Hinton, et al., Managing extreme AI risks amid rapid progress (May 2024).
We helped organize a consultation on 'AI -- Threats and Opportunities' at St George's House, Windsor Castle in October 2023, at which some of these issues were discussed. The report of the consultation is available here.