Hello again, Tech Brief readers! It’s 40 days out from the election — and we’re once again watching upheaval at OpenAI, while your favorite actors are lending their voices to Meta’s AI. I’m Cat Zakrzewski, filling in for Cristiano. Got news tips? Send them to: [email protected].
Below: A new report finds U.S. schools are rife with nonconsensual intimate imagery. But first, let’s fast forward to 2025 and talk about France.
France gets set to host a very different kind of AI summit
At the peak of the AI hype cycle last November, governments from six continents and the world’s most famous tech executives descended on the symbolic birthplace of the digital era for the first global artificial intelligence summit.
The Bletchley Park Summit in Britain drew Vice President Kamala Harris, OpenAI CEO Sam Altman, Tesla CEO Elon Musk and more among the who’s who of the world’s most powerful government officials and tech executives. The summit produced an international agreement with buy-in from China to contain the most extreme AI risks. But it was widely panned for focusing too much on the doomsday scenarios and excluding voices about the technology’s impact on workers or the Global South.
Almost a year later, France is gearing up for its chance to host the global event, dubbed the AI Action Summit. And it’s going to be very different.
Anne Bouverot, France’s special envoy for AI, says this summit is going to be bigger, more inclusive and more focused on ways to steer the technology to serve the public interest. Rather than emphasize frontier risks, the French event will address how AI is changing work, supercharging disinformation and impacting the climate and health.
“We’re going to focus on concrete action and deliverables,” Bouverot told The Washington Post. “We’re not here to reinvent global governance.”
This week, Bouverot attended AI events during the United Nations General Assembly, which brought Altman and other international leaders to New York for discussions about AI policy. The French summit will build on the work the U.N. and the Bletchley Park Summit have done to coordinate global AI regulation, she said.
France’s ambitions to have an action-focused summit are likely to come up against the challenges of global technology governance, which often relies on voluntary and nonbinding commitments from nations and companies.
AI and the companies that produce it are not bound by state borders but governed by a patchwork of different regimes. The European Union, which includes France, is adopting an AI Act, which seeks to ban unacceptable uses of the technology, like social scoring systems, and restrict other uses based on a sliding scale of risks. China has embraced rules that promote its socialist values and censorship. The United States has largely left AI regulation up to the states, which are weighing measures to limit deepfakes, protect elections and create greater liability for AI companies.
Bouverot said France will address this challenge by expanding existing international AI partnerships. For instance, the event in France could build on the recently launched international network of AI safety institutes to evaluate AI systems, she said.
“We’re not going to do this alone,” she said.
While the location and attendees have not yet been announced, Bouverot said many tech executives from large companies have accepted Save the Dates. France is trying to expand the guest list beyond industry, as well, and the Paris Peace Forum is soliciting applications for initiatives that promote AI for the common good.
For its part, U.S. involvement in the summit hinges on the outcome of the presidential election.
The summit is scheduled for Feb. 10 and 11, less than a month after the inauguration. The initiatives reflect many of the values the Biden-Harris administration has promoted in its AI executive order, and Harris was the face of many of those efforts on the international stage last year.
However, if Donald Trump is elected, it’s likely his “America First” agenda would be at odds with the goals of the summit. Trump has promised to overturn Biden’s AI executive order and replace it with his own. His allies have circulated a framework that includes a section called “Make America First in AI.” Republicans have criticized efforts to address disinformation as censorship, and the summit’s focus on climate may also create divisions with a potential Trump administration. During Trump’s term, he withdrew from the Paris climate accord.
Bouverot declined to comment on how a potential Trump presidency would impact the summit.
“Of course, AI is political,” she said. “But it’s something that’s so broad, that’s coming so quickly, that’s having an impact on so many things. Every government is very much trying to figure things out.”
From our notebooks
‘Tech-powered sexual harassment’ pervasive in schools, study finds
U.S. schools are rife with nonconsensual intimate imagery, sometimes referred to as “revenge porn,” according to a new survey of students and teachers conducted this summer by a tech nonprofit, my colleague Cristiano Lima-Strong reports for the Tech Brief.
Participating students and teachers reported observing “substantial amounts” of such material in grade schools, including human-made content and imagery developed with artificial intelligence, the Center for Democracy & Technology (CDT) study released Thursday found.
The survey also highlighted potential hurdles to tackling the issue. Teachers report that very few schools have proactive policies, and when they do respond, they often focus on punishing the perpetrators but not aiding victims. Female and LGBTQ+ students were the most likely to be impacted, students and teachers reported.
The issue has been a growing point of concern for legislators and regulators amid the rise of AI and could come into greater focus under a potential Kamala Harris administration. While serving as California’s attorney general, Harris made curbing such material a signature issue.
The nonprofit organization, which received funding from tech companies and philanthropic groups, conducted its survey online between June and August with 1,316 ninth- to 12th-grade students, 1,006 sixth- to 12th-grade teachers and 1,028 sixth- to 12th-grade parents.
Government scanner
Inside the industry
Competition watch
Trending
Daybook
- Public Knowledge hosts the “IP3 Awards,” Thursday at 6:30 p.m.
- Finra holds an event, the “Finra Advertising Regulation Conference,” Thursday and Friday.
- Georgetown Law holds an event, “AI on Trial: Liability in the AI Ecosystem,” Friday at 9:15 a.m.
Before you log off
That’s all for today — thank you so much for joining us! Make sure to tell others to subscribe to Tech Brief. Get in touch with Cristiano (via email or social media) and Will (via email or social media) for tips, feedback or greetings!