Why it matters: As AI tech accelerates, the Biden administration is intensifying its focus on its risks and regulation, pushing industry heavyweights to pledge proactive security measures. This represents one of the most significant actions taken by the administration to address potential threats posed by AI.
The big picture: Seven major companies, including Google, Amazon, Microsoft, Meta, and OpenAI, have committed to pre-release security tests by independent experts, data sharing on system safety with the government and academia, and the creation of alert systems (“watermarking”) to identify AI-generated content. AI newcomers Anthropic and Inflection have also signed the pledge.
The stakes: The FTC could penalize companies for deceptive practices if they fail to uphold their commitments. This initiative is a part of Biden’s broader strategy to raise safety, security, and trust standards across the AI industry.
What’s next: This is only the beginning of the administration’s work on AI. An AI-focused executive order is in the works, and bipartisan legislation to regulate AI technology is being supported. High-level executives from the pledge-participating companies, including Microsoft President Brad Smith, Inflection AI CEO Mustafa Suleyman, and Meta President of Global Affairs Nick Clegg, are set to meet with the White House.
But, but, but: Despite the commitments, critics are wary due to tech companies’ spotty record in following through with such pledges. Nonetheless, the companies have reiterated their commitment to advancing responsible AI.
The global context: The EU is also making strides in AI regulation and is currently negotiating its own EU AI Act, encouraging tech firms to make similar commitments through their “AI Pact”.