Board Politics and Anthropic’s Unique Mission

If you’ve never heard of Anthropic’s Long-Term Benefit Trust (LTBT), this arm of the firm is supposed to be a counterweight to investors, a check and balance on unfettered development that doesn’t align with the core mission of keeping AI safe. That’s what the “long-term” part means: the long-term benefit of humanity – not just profit.

How does Narasimhan fit in?

Well, there are a few things in play. Over here, it looks like Anthropic is also maybe looking to do an IPO. But there’s also a shift that changes the dynamic of the board in terms of the LTBT.

Here’s how it’s put at an information resource at AIProductivity that draws straight from Anthropic’s own blog:

“For the first time, Anthropic's independent oversight body controls the majority of the company's board. The Long-Term Benefit Trust (LTBT) - a group of directors with no financial stake in Anthropic, designed to keep the company accountable to its stated mission - has appointed Novartis CEO Vas Narasimhan as a new board member, pushing Trust-appointed directors past the 50% threshold.”

Neil Shah, the LTBT chair, had this to say in a statement:

"The Long-Term Benefit Trust's role is to appoint directors who will ensure Anthropic responsibly balances its commitment to stockholders and its public benefit mission as the company grows. Vas has spent his career stewarding breakthrough science responsibly —exactly the perspective we are excited to have on the board as we develop consequential technology. We’re excited for what he’ll bring to the table.”

Narasimhan made his own statement on the appointment, too:

“Working across medicine, innovation, and global health has shown me the transformative potential of technology when deployed responsibly,” he said. “In healthcare, AI is accelerating solutions to some of the hardest scientific challenges, from deepening our understanding of disease biology to designing better medicines. Anthropic is setting the standard for how AI should be developed to benefit humanity, and I’m honored to join the Board and contribute to its mission.”

So it turns out this isn’t just another board seat – given everything that’s been in the news, and Anthropic’s reputation as a company that wants to be safe, it’s a big deal. Some of the reporting notes that having an LTBT or anything like that is extremely unusual in Silicon Valley, and in AI.

Other Developments

Of course, as a front-runner in the industry, Anthropic is also moving aggressively in hardware. I reported a week ago about the agreement with Google and Broadcom for TPUs (instead of GPUs), and other developments also support the company’s clear goals, as well as the track record that’s constantly evolving.

And then there’s Claude’s marks on benchmarks. SWE-bench, GPQA, and more. Mythos is changing the game, and people are noticing, including the Fed chair and banking bigwigs, who met recently to discuss how the model might compromise banking security.

Sometimes I wonder what the old guard at MIT would have thought, if they could see what’s happening now. After all, Claude Shannon is the Anthropic model’s namesake, and himself and Marvon Minsky and a few others are widely regarded as the “fathers” of AI. Would they be shocked at how things are progressing? Probably.

Anyway, look for Anthropic to make further inroads into healthcare. And as for the LTBT: we’ll see if having the majority helps to keep Claude on the rails when it comes to safe and secure operations.