Godfathers of AI warn developers of the technology must be held liable

A group of 23 senior experts, including two of the three "godfathers" of artificial intelligence (AI), who won the ACM Turing award in 2018, have warned that powerful AI systems threaten social stability and AI companies should be held liable for any harm caused by their products. The warning was issued on Tuesday ahead of an AI summit being held at Bletchley Park next week. The experts urged governments to pursue policies such as allocating one-third of AI research and development funding to safe and ethical use of systems, giving independent auditors access to AI laboratories, and establishing a licensing system for building cutting-edge models. The experts also called for AI companies to adopt specific safety measures and be held liable for foreseeable and preventable harms from their AI systems.

In a related article, the CEO of Google's AI unit in the UK, Demis Hassabis, called for the world to treat the risks posed by AI as seriously as the climate crisis. He warned that the world cannot afford to delay its response to the dangers of AI and that oversight of the industry should start with a body similar to the Intergovernmental Panel on Climate Change (IPCC). Hassabis believes that a regime of oversight is needed, and governments should take inspiration from international structures such as the IPCC. He has suggested that an equivalent of a Cern for AI safety be established for international research and, eventually, an equivalent of the UN's International Atomic Energy Agency (IAEA), which actually audits these things.