Thanks for reading The Briefing, our nightly column where we break down the day’s news. If you like what you see, I encourage you to subscribe to our reporting here.
Greetings!
Well, that didn’t take long. Jensen Huang’s time as CEO of Nvidia appears to be coming to an end, according to multiple reports that the board was meeting Wednesday night to consider his future. Just three days ago, it announced an inquiry into whether he was the “target of a foreign intelligence effort,” but at the time the board said it had “confidence in Jensen.” It seems likely that confidence was undermined by a New York Times story on Wednesday that said Huang had had an affair with a company executive in addition to an industry executive who may also be a Chinese spy.
If the board dumps Huang, it will demonstrate that no CEO is indispensable, no matter how vital their role is. Many observers had insisted the opposite in recent days, perhaps understandably. After all, Huang co-founded the company and still owned 3.6% of the stock, making him one of the biggest shareholders. As we wrote in this piece, he “has a hand in nearly every aspect of the company’s day-to-day operations.” Along with Sam Altman, he’s been probably the most articulate advocate for AI. Most importantly, his vision of the future has positioned Nvidia to benefit more than any other company from the AI boom, driving its market capitalization to a high of $5.8 trillion in early October, according to Koyfin.
Not surprisingly, investors have freaked out about the possibility of his exit. After falling 10% in the first two days of this week, Nvidia was trading down 8% in after hours trading on Wednesday to around $160, giving the company a market capitalization of around $3.85 trillion. The question now is who will replace him. There are no obvious internal candidates, although Nvidia could appoint an executive as interim chief.
That suggests the board will have to go outside Nvidia. Earlier this week we suggested that either AMD CEO Lisa Su or Arm CEO Rene Haas could succeed Huang. On TITV on Tuesday, D.A. Davidson analyst Gil Luria suggested Intel CEO Lip-Bu Tan. This is a big job. Who can do it?
Anthropic’s Not-So-Powerful Trust
Anthropic executives are meeting with investors this week, ahead of next week’s expected IPO. Here’s one thing investors should ask: How concerned should they be about the Long-Term Benefit Trust, a secretive group that has the power to appoint and remove a majority of Anthropic’s board? Their job is to ensure the company is run in a way that balances the financial interests of shareholders with Anthropic’s mission to develop advanced AI “for the long-term benefit of humanity.” What a balancing act!
And who controls the trust? It turns out that the trustees—who include a former Fed chair, a health policy wonk, and a physics professor at Harvard University—pick future trustees. (There’s a vacancy for a fifth trustee, by the way, and it pays $400,000 a year!) In other words, it’s a closed club, which seems to give those four an extraordinary amount of power. But things are not exactly what they seem. A big restriction on the trust’s power essentially defangs it.
The trust appoints and can in theory remove four of Anthropic’s seven directors. In reality, it can only remove three of those four unilaterally. A committee of Anthropic’s board, which includes Anthropic co-founder Daniela Amodei, has to approve the trust’s removal of the fourth board member. If the board committee opposes the director’s removal, Anthropic’s employees get involved and have to approve it. As Amodei and other company co-founders essentially pick three of the seven board members, the co-founders could counter any effort by the trust to replace a majority of the board.
The board committee also has a role if the trustees want to remove one of their members. Anthropic says in its filing that “these procedural requirements are intended to promote collaboration between the Trustees and our leadership in identifying and sourcing candidates” for the board and the trust. The rules “do not give our board of directors the substantive ability to limit the LTBT’s freedom to elect directors of its choosing.” It’s hard to take that statement seriously, given that it’s precisely what the requirements do.
The point is that it’s the founders rather than the LTBT who control the company. For investors, that’s probably good news, given how aggressively Anthropic competes—see here and here—and given its success in overtaking OpenAI. Of course, as AI models become more sophisticated and more powerful, a clash over its risks to society seems likely. Don’t be surprised if one day a huge fight breaks out between the trust and the board reminiscent of OpenAI’s 2023 brawl over the short-lived removal of CEO Sam Altman. When that happens, expect the founders to prevail.
The Information’s Top 50
If there’s one field every parent should encourage their child to go into, it’s cybersecurity. The spread of AI, with the ability to uncover holes in software that hackers can exploit, is becoming a huge deal. NBC News reported this week that cybersecurity firms were “turning away new clients” because they were overwhelmed!
It shouldn’t be surprising, then, that cybersecurity is a hot sector for startups. In our latest installment of The Information 50—our annual list of the most promising startups out there—three of the four most highly valued companies are security startups. Here’s the list.
In Other News
• Meta Platforms is changing the leadership of its massive data-center operation, with longtime executive Santosh Janardhan telling employees that he is stepping down as head of global infrastructure and will leave the company early next year.
• United Airlines said Wednesday it will ban AI agents and assistants from booking its flights, following similar bans by banks and retailers that cited privacy concerns.
Today on The Information’s TITV
Check out today’s episode of TITV in which Akash Pasricha interviews Adobe executives at their Max conference about standing out in the crowded market of AI creative tools.