OpenAI isn’t being consistently candid about Leading the Future
OpenAI says it doesn’t fund or direct LTF — but one of the super PAC’s operatives has described it as a “corporate funder” with “a say”
When OpenAI published a post on June 1 distancing itself from the super PAC Leading the Future, employees were rather pleased.
In recent weeks, OpenAI employees have raised concerns to the company’s global affairs team about OpenAI’s links to the PAC, to which its president Greg Brockman and his wife Anna have donated $25m, as well as some of OpenAI’s own policy actions, one employee told Transformer. At a tense meeting in May with global affairs chief Chris Lehane, they pressed OpenAI to clarify its relationship with the super PAC, the employee said.
Shortly afterwards, the company publicly addressed concerns over LTF, stating, among other things, that: “OpenAI does not direct the activities of LTF, or have visibility into their operations.”
The blog appeared to reassure some employees. “I’m happy OpenAI put out this statement,” alignment researcher Jason Wolfe said, noting that “personally I really dislike a lot of things I’ve heard about LTF.”
“The good news is that we can now show congressional staff this blogpost that says in bold ‘No outside political group speaks for OpenAI or represents our company’s views.’ and criticizes LTF tactics,” Shantanu Jain, another technical researcher at the company, said, quoting the post.
But the post does not tell the full story, omitting key details about Lehane’s role in establishing LTF and the way super PAC operation’s own staff view their relationship with OpenAI.
As previously reported, Lehane was one of a group who provided advice on creating Leading the Future, its affiliated partisan super PACs, and its affiliated 501(c)(4) Build American AI, which does not have to disclose its donors.
What has not been reported is that he is thought in AI policy circles to have picked Josh Vlasto, a longtime Democratic operative, to co-lead LTF — something that when asked, neither OpenAI nor Leading the Future denied. And though OpenAI has insisted that Brockman’s donation was made in a “personal capacity,” Transformer has found that not everyone working within LTF appears to view it that way.
“For this project I’m funded by four separate entities … OpenAI is just one of them,” Nathan Leamer, executive director of Build American AI, told Transformer over text message on May 3, just a month before OpenAI’s statement.
When later asked about the division of labor between Leamer, Moffat, Vlasto and Lehane, Leamer said “I think the best way to think of it is the corporate funders all have a say but they interact directly with Zac and Josh. They run LTF which funds my crew BAAI.”
In response to a request for comment, LTF spokesperson Jesse Hunt told Transformer: “Neither Leading the Future nor any affiliated super PACs or 501(c)(4) organizations has ever received funding from OpenAI. All decision making regarding candidates, policy or operations are made solely by Zac Moffat and Josh Vlasto.”
An OpenAI spokesperson said: “Neither OpenAI nor Greg and Anna Brockman have donated to or been involved with Build American AI.”
She added: “As Chris has said previously, he was consulted when the PAC was getting stood up last year, but has no current involvement.”
In recent weeks, Lehane has spoken to reporters at WP Intelligence, Wired and Punchbowl. In those interviews, Lehane said he was displeased with some of Leading the Future’s activities, telling WP Intelligence that OpenAI wasn’t “so much into the tactics.” He also emphasized the separation between OpenAI and Leading the Future’s donors, telling Wired that he had only advised Brockman “in a very general way” over his political spending, and saying that he is currently “not involved” with Leading the Future and wants to “let them be their own independent outside thing.” Earlier in the month, an OpenAI spokesperson said that OpenAI had “not provided funding or any support to” Leading the Future or Build American AI. That appears to directly contradict Leamer’s inclusion of OpenAI as one of the group’s “corporate funders.”
In the June 1 post, OpenAI also set out a set of standards for lobbying and political advocacy on AI, writing that groups “that are advocating on AI should … not use tactics like astroturfing that obscure the real choices facing policymakers and the public.” Days later, Build American AI admitted that one of its “outside vendor[s]” was behind multiple anonymous sock puppet X accounts exposed by the Midas Project and Taylor Lorenz. When Lorenz previously uncovered an LTF-linked campaign to pay influencers to spread anti-China messaging, she reported that “the agency hired to execute it said they reported directly to Vlasto.”
Ironically, given the lengths OpenAI has gone to state that it has no sway over LTF, recent changes in the super PAC’s approach appear to align with OpenAI’s own policy shifts in a way that many of its more safety-minded employees might welcome. In the weeks since OpenAI employees expressed disappointment about its support for a controversial Illinois bill that would have shielded AI companies from liability, OpenAI dropped its support for the bill, instead endorsing a much stronger bill, SB 315. OpenAI suggested that bill, along with New York’s RAISE Act and California’s SB 53, could set the standard for federal laws in what Lehane has dubbed “reverse federalism.”
In the same period of time, Build American AI and LTF have demonstrated a similar shift in messaging. Shortly after OpenAI endorsed SB 315 (and after the bill passed), Build American AI also endorsed the bill. Leading the Future, meanwhile, recently endorsed the RAISE Act, despite spending millions opposing its sponsor, Alex Bores, and criticizing a “patchwork of regulation.” It is a far cry from last year, when Leading the Future and Build American AI ran a $10m campaign calling for federal preemption of state AI laws.
OpenAI is of course not the group’s sole influence. The co-founders of venture capital firm Andreessen Horowitz (a16z), Marc Andreessen and Ben Horowitz, have each contributed $25m to the super PAC, and Leading the Future has often taken positions similar to those of the VC firm. Like LTF, a16z has long argued against legislating for AI at the state level, warning of a “patchwork of state laws.” The American Innovators Network, another a16z-backed group, opposed SB 315 and the RAISE Act. A16z declined to comment.
With a16z and OpenAI’s stances seemingly diverging, the pressure to appease both parties puts the super PAC and Build American AI in a tricky situation. The group has produced friendlier messaging on some state AI laws which OpenAI supports, while continuing to spend money backing political candidates opposed to stringent AI regulation.
With each step towards stronger AI safety standards, OpenAI shifts further from a16z, creating a tension for LTF as the midterms approach. It and its affiliates have so far deployed more than $18m on campaign ads, with tens of millions remaining in their war chest.
Additional reporting by Shakeel Hashim.