The White House brings together seven of the largest AI companies to make a series of voluntary commitments to protect users.
Companies comprising Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI have agreed to respond to a series of requests from the White House to discuss many of the risks posed by artificial intelligence. Promises to be made include investments in cybersecurity, discrimination research, and a new watermarking system that notifies users when content is created by artificial intelligence.
Since companies voluntarily participate in these agreements, they will not face any sanctions for now if they do not fulfill their commitments. Many of these commitments are not expected to take effect immediately, but companies are requested to implement them at the earliest opportunity.
A White House official told reporters on Thursday that the Biden administration is currently working on a number of laws to address some of the risks posed by artificial intelligence.
The Biden administration has met with tech executives, labor and civil rights leaders over the past few months to discuss AI. In May, the White House announced more funding and policy guidance for companies developing AI technology, including $140 million to the National Science Foundation to establish seven new National Artificial Intelligence Research (NAIR) Institutes. Google, Microsoft, Nvidia, OpenAI and other companies have also agreed to allow language models to be publicly evaluated at this year’s Def Con.
The announcement comes nearly a month after Senate Majority Leader Chuck Schumer launched a plan for Congress to regulate technology without impeding innovation. The plan, called the SAFE (Security, Accountability, Fundamental principles, Disclosure) Framework, doesn’t make specific policy demands, but urges lawmakers to work together to create rules that address the potential of AI to harm national security, cause job loss, and create misinformation.
In addition, Schumer’s plan includes a series of information about technology for senators. The third information session will take place next week. The first two meetings explained the technology and presented its risks to national security.
Since Schumer’s announcement, lawmakers on both sides have introduced legislation to regulate the technology. Some new rules limiting how the Department of Defense can use manufacturer artificial intelligence have made their way into the National Defense Authorization Act, which should have been mandated this year. Senators are expected to vote next week to approve the measure, CNN reported.
Representatives from all seven companies are expected to meet today at a summit at the White House to physically sign these commitments.
55928
