Leading AI Developers Agree to White House’s AI Safety Principles
Seven major U.S. AI companies have agreed to a set of safety and responsibility principles, announced by President Biden at the White House. Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI are all committed to addressing growing concerns around generative AI, synthetic media, and deepfake technology, including independent audits and digital watermarking.
The broad agreement covers much of how generative AI has been utilized as well as possible future misuse. The companies all agreed to extra security tests of their products and more cybersecurity investment. Additionally, the companies have promised to publicly disclose information about flaws and risks discovered in their AI technologies, including potential issues related to fairness, safety, and bias. On the deepfake side, the companies have agreed to work on ways to detect when deepfakes are deployed. That’s something most of the companies involved have already begun developing, at least for visual and audio content. The pledges are short on technical details but include commitments to use AI “to help address society’s greatest challenges, citing medical and environmental research as examples of how it might be employed.
“Artificial intelligence or — it promises an enormous — an enormous promise of both risk to our society and our economy and our national security, but also incredible opportunities — incredible opportunities,” President Biden said in a short speech following the signing. “These commitments are real, and they’re concrete. They’re going to help fulfill — the industry fulfill its fundamental obligation to Americans to develop safe, secure, and trustworthy technologies that benefit society and uphold our values and our shared values. Let me close with this. We’ll see more technology change in the next 10 years, or even in the next few years, than we’ve seen in the last 50 years. That has been an astounding revelation to me, quite frankly. Artificial intelligence is going to transform the lives of people around the world.”
The White House issued a statement outlining the agreement but described the voluntary measures as a temporary bridge for the industry to carry them into a new regulatory environment once Congress passes legislation to regulate AI. Senate Majority Leader Chuck Schumer (D-NY) has said he plans to introduce a bipartisan bill on AI governance in the near future, something OpenAI CEO Sam Altman encouraged at a U.S. Senate hearing. The European Union has its own upcoming regulation proposals, but the U.S. and the rest of the G7 announced plans to devise an international standard for generative AI, with the “Hiroshima AI Process” producing a report by the end of the year.