Anthropic's IPO filing this week contained myriad revelations about the state of the AI firm's finances, as well as an eye-opening series of claims about its commitment to helping others. We asked three experts in the world of philanthropy what they thought of these giving pledges. Their verdict: we'll see.

The company behind the AI chatbot Claude, which filed this week to go public in what could be the biggest IPO in history, laid out how it plans to give back and encourage its employees to do the same. Anthropic's seven cofounders said they've pledged 80% of their personal stock in the startup to charitable causes, gifts that could together be worth $80 billion. They also said they've earmarked 5% of Anthropic's equity for a "fund devoted to the public good," and plan to set aside another 2% for every sustained 50% rise in the company's stock price, with its IPO price as the base. Moreover, they said that new employees are offered a donation match, meaning that for every share they pledge to give away to good causes, Anthropic creates an additional share and matches the donation.

Anthropic said that just over half of its employees have pledged at least some of their equity, including 70% of employees hired before May 2024, and nearly 50% of those hired after that date. The company noted that 25% of the former group, who were offered up to a 50% match, pledged the full amount. That figure was nearly 30% for the latter group, who were offered a 25% match.

Anthropic also said it plans to issue Class E shares — which carry limited voting rights but have no economic value — to reward tenure and charity. Employees stand to receive one share for every six months they stay at the company, and one Class E share for every normal share they pledge to donate, Anthropic said. Anthropic didn't respond to a request for comment from Business Insider.

Three philanthropy experts told Business Insider via email this week that Anthropic's giving plan is promising but represents only a first step.

"It's admirable, and it's well-designed," said Dr. Vandana Arcot, the founder and CEO of Komselj, a philanthropic strategy firm. She noted that "building it into the compensation structure makes generosity the default, not the exception." Arcot said Anthropic's giving seemed to be grounded in an "ethic of accountability, not abundance." She pointed to CEO Dario Amodei's January essay in which he said that people who get rich from AI have a responsibility to address its economic fallout.

Arcot also cited Anthropic's $200 million commitment this summer to fund research on preparing people for AI's economic impacts. She said that valuing generosity "tends to help companies attract mission-driven talent and keep it, which is a business case on top of the moral one." She added that "defining your giving in advance is a way of saying: judge us on this." But Arcot said her full evaluation of Anthropic's efforts will depend on "who sits on the decision-making body, how recipients are chosen, and what gets reported."

Bella DeVaan, the director of the Charity Reform Initiative and deputy director of the Program on Inequality and the Common Good at the Institute for Policy Studies, said there's such "dire" wealth inequality in the US that "any efforts to center the public benefit are commendable." However, she said they're "no substitute for structural reform," and "giving to the public benefit shouldn't be voluntary, but should be guaranteed by our tax system." DeVaan also noted the charitable pledges by Anthropic's cofounders aren't legally binding and can be revoked.

She said it was "concerning" to see that personal interest could trump public benefit, and added that it's "exactly why voluntary efforts alone won't be sufficient to rely on."

Hans Peter Schmitz, a professor of nonprofit leadership at North Carolina State University, said he applauded Anthropic for promoting generosity from the top down, but it struck him as "part of an intentional effort in managing public opinion." He said that any positive publicity around Anthropic's giving could help differentiate it from competitors and counter bad news about AI. The nascent technology has proven contentious due to concerns about job losses, environmental damage, cybersecurity threats, and even danger to humans.

Schmitz said Anthropic appeared to be engaging in "reputational management," as "pledges are a relatively cheap 'signal' to be viewed in a positive light." "Like many pledges in the past, there is also a high likelihood that it may never happen and/or a lot of the funding is channeled into a narrow set of issues, like AI safety," he said.