The White House’s latest plans to keep artificial intelligence tools in check include pressuring tech leaders, funding more research centers, and supporting the world’s largest AI hackathon at the upcoming DEFCON 31 cybersecurity conference.
Vice President Kamala Harris hosted a meeting Thursday with the heads of four of the companies heavily invested in AI to “share concerns” about the risks associated with the technology.
In attendance was OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Microsoft Chairman and CEO Satya Nadella, and Google and Alphabet CEO Sundar Pichai.
President Joe Biden joined the meeting briefly, telling the industry leaders: “What you’re doing has enormous potential and enormous danger.”
The meeting was touted as part of a broader ongoing effort by the White House to engage with advocates, companies, researchers, civil rights organizations, not-for-profit organizations, communities, international partners and others on critical AI issues.
“The private sector has an ethical, moral and legal responsibility to ensure the safety and security of their products,” Harris said in a statement after the meeting.
According to a White House readout of the meeting, it included “frank and constructive discussion” on three key areas: the need for companies to be more transparent with policymakers, the public, and others about their AI systems; the importance of being able to evaluate, verify, and validate the safety, security, and efficacy of AI systems; and the need to ensure AI systems are secure from malicious actors and attacks.
The non-attendance of Facebook founder and Meta CEO Mark Zuckerberg surprised some industry observers, given the company’s heavy investment in AI.
An administration official told CNN Meta was not invited to the meeting, because it was “focused on companies currently leading in the space, especially on the consumer-facing product side.”
Also on Thursday, the Administration announced new initiatives to promote responsible AI innovation.
The National Science Foundation will spend $140 million to fund the launch of seven new National AI Research Institutes, bringing the total number of institutes to 25 across the country. The Institutes are intended to operate as a catalyst for collaborative efforts across institutions of higher education, federal agencies, industry, and others to pursue transformative AI advances that are ethical, trustworthy, responsible, and serve the public good.
The White House will also support what is being described as the largest-ever public generative AI red-teaming event, to be held at DEFCON 31 in August.
AI developers Anthropic, Google, Hugging Face, Microsoft, NVIDIA, OpenAI, and Stability AI have agreed to participate in the hackathon which will see thousands of DEFCON attendees hunting for bugs in their large language models.
“This independent exercise will provide critical information to researchers and the public about the impacts of these models, and will enable AI companies and developers to take steps to fix issues found in those models,” the White House said.
Sven Cattell, the founder of AI Village which is hosting the hackathon at DEFCON, said the “diverse issues” associated with large language models would not be resolved until more people knew to assess and de-bug them.
“Bug bounties, live hacking events, and other standard community engagements in security can be modified for machine learning model based systems. These fill two needs with one deed, addressing the harms and growing the community of researchers that know how to help.”
The White House said its newly-announced efforts to ensure AI innovation occurred responsibly followed earlier initiatives including its Blueprint for an AI Bill of Rights and related executive actions announced last fall, the AI Risk Management Framework and a roadmap for standing up a National AI Research Resource released earlier this year.