The Tech Company Stocking Up On Democrats As Silicon Valley Turns Right

SAN FRANCISCO — Sam Altman, the driving force behind ChatGPT’s meteoric rise, is running a team of veteran political operatives, campaigning to secure his company OpenAI’s future. Only in this case, there’s no gray-at-the-temples candidate.

Instead, the billionaire CEO is in an existential race to remain at the top of the hypercompetitive artificial intelligence market, with rivals like Google, Meta, Anthropic and Elon Musk’s xAI gunning for the lead, at times dangling nine figures to poach top talent.

Over the past year alone, the world’s most closely watched AI company has hired more than half a dozen political insiders who are well-connected to the Democratic establishment, from Bill Clinton’s former spin doctor Chris Lehane to Kamala Harris’ one-time bestie Debbie Mesloh and ex-Sen. Laphonza Butler.

It’s a notable deviation at a time when much of Silicon Valley is more focused on staffing up to chase influence in Republican-controlled Washington. And it’s among the most aggressive pushes to date from a tech company into Sacramento and other corridors of power in a state that birthed the industry, yet where firms had long been reluctant to engage directly at the levels of other major sectors.

But it underscores how OpenAI sees its deep-blue home of California as vital for its global ambitions — tied to a planned business makeover that the state’s top attorney can summarily shut down.

“They’re bringing in some very big guns to make their case,” said Orson Aguilar, president of the nonprofit LatinoProsperity and prominent critic of OpenAI’s business transformation plans within the state.

“Since the stakes are so high here for their profit, they’re willing to spend what it takes to get their way with the California attorney general,” added Aguilar, a seasoned civic leader.

POLITICO interviewed two dozen people who have interacted directly with the company, worked with its new hires or demanded answers about OpenAI’s business moves.

The conversations revealed how recent recruits have drawn on tactics from the campaign trail and from warding off political scandals; from raising doubts about critics to burnishing the company’s public perception through links to respected California figures such as Dolores Huerta, who co-founded the United Farm Workers with César Chávez.

At the heart of their campaign is OpenAI’s bid to change its business model, which is facing a lawsuit from Musk — the company’s co-founder turned rival — as well as an investigation by California Attorney General Rob Bonta. Central to the approach is sniffing out any potential whiff of Musk — a divisive figure to Californians and the omnipresent boogeyman in OpenAI's righteous, dare they contend, underdog quest — when new criticisms arise, POLITICO’s reporting shows.

Founded as a nonprofit, OpenAI views the restructuring — which could attract hundreds of billions of dollars from investors — as the only way to keep up with the industry’s global cash dash. It says the changes are necessary to continue the nonprofit’s founding mission of ensuring human-like artificial intelligence benefits all.

But that new structure has to first clear the circling Democratic regulators and intensifying public scrutiny in California. The hires have been messaging that the company can still be a force for good, while defusing backlash from skeptics who argue OpenAI has put profit over mission.

They’ve also sought to make the stakes feel just as high for California, which depends on the tech industry’s tax revenue to fill its coffers and has already seen some big companies, including those run by Musk, leave amid clashes with regulators and complaints about an unfriendly business climate.

“That’s a question that folks should be thinking about because I do think that we want to be here,” said Lehane — now OpenAI’s chief global affairs officer — in an interview. “So I’m hoping folks make the right decision.”

The players

The campaign manager of OpenAI’s political strategy has been Lehane, who earned the nickname “master of disaster” during his Clinton days, handling the Monica Lewinsky affair and other scandals. He literally wrote the guidebook for politicians on damage control, and was also Al Gore’s press secretary.

Known for his sometimes theatrical strokes and quiet, often-unreported counseling sessions with bold-faced names throughout Democratic politics, he later guided billionaire climate crusader Tom Steyer’s deep-pocketed electoral efforts (including his flirtation with a 2016 U.S. Senate run ultimately undertaken and won by Harris) before making a jump to the tech world with Airbnb in 2015.

The San Francisco resident joined OpenAI last summer in the aftermath of internal upheaval that rocked the company: a boardroom coup that briefly removed Altman as CEO.

Lehane told POLITICO that dating back to his time leading global policy at Airbnb, he’s modeled his operations off a political campaign “in the sense that you really want to integrate a public narrative with real, substantive ideas.”

But not as in “running negative political ads,” he was quick to clarify.

Lehane’s campaign analogy extends to how he has organized the new recruits, dividing them between communications, policy and field assignments. (The opposition researchers, be they in-house talent or hired guns, have thus far managed to remain anonymous.)

Since Lehane took over, OpenAI’s global affairs team has staffed up with Gavin Newsom and Bill de Blasio alum Peter Ragone; Harris’ old campaign confidant and friend Mesloh; and Marisa Moret, who was chief of staff for the San Francisco city attorney while Harris was district attorney. Moret was Lehane’s right hand at Airbnb, and reports to him again at OpenAI.

Ragone, a bicoastal fast-talker who revels in piecing together complex political relationships by closely following contribution breadcrumbs, is a damage control expert himself who met Lehane during the 1990s. In the New York mayor’s office, he attempted to smooth over public perception amid de Blasio’s rift with a prominent police union following the 2014 shooting of two officers.

While Newsom was mayor of San Francisco, Ragone helped the current governor navigate an affair he had with the wife of his campaign manager. Recently, he worked on billionaire shopping mall magnate Rick Caruso’s unsuccessful run for Los Angeles mayor and is a confidant of Lt. Gov. Eleni Kounalakis.

Ragone’s interest in AI and tech policy can be traced back years. He circulated memos to Newsom and other policymakers on the topic well before signing on with OpenAI.

Ann O’Leary, another Newsom alum who served as the governor’s first chief of staff and co-directed Hillary Clinton’s 2016 presidential transition team, is spearheading some of OpenAI’s recent legal activities.

Now an attorney at Jenner & Block, she’s at the center of the company’s dealings with investigations by both California’s attorney general, Bonta, and Kathleen Jennings, attorney general for Delaware, where OpenAI is incorporated. O’Leary’s role has also involved probing opponents and drawing parallels to Musk.

Lehane has additionally turned to outside consultants, namely former Harris campaign manager Brian Brokaw and his business partner Dan Newman, for help crafting OpenAI’s public image.

The two longtime Newsom advisers have a track record of blending campaign chops with insider relationships in Sacramento and the Bay Area to try to move other controversial initiatives forward, such as the billionaire-backed California Forever project to build a new tech world utopia north of San Francisco.

At the same time, they’ve managed to stay close with multiple top officials, from Bonta to San Francisco Mayor Daniel Lurie, while quietly assembling a portfolio of corporations in crisis. Brokaw and Newman formed a pro-Lurie committee to raise money for his successful City Hall bid.

OpenAI brought on multiple hires with tight ties to Harris when she was still considered a potentially field-clearing contender in the 2026 governor’s race and even in the midst of her presidential run last year. Harris only ruled out a run to replace Newsom in late July.

Butler, whom Newsom personally hand-picked to fill the late California Sen. Dianne Feinstein’s seat, is advising OpenAI as well, POLITICO first reported. Before the Senate, Butler was the longtime leader of a prominent labor group in California and a campaign adviser to Harris via the same political firm where Newman used to work. Butler’s previous consulting clients include ride-hailing giant Uber and Airbnb, where she overlapped with Lehane and Moret.

An OpenAI spokesperson said the restructuring was among the issues Butler works on. The company declined requests to interview other members of Lehane’s team.

Neither Newsom’s office nor Harris’ representatives responded to requests for comment on their former staff.

The playbook

Lehane is credited with teaching Silicon Valley how to play politics, first at Airbnb, then at cryptocurrency exchange Coinbase, where he was an architect of its emergence as an election-swaying force.

At OpenAI, he has to move differently. The company’s nonprofit status prevents it from launching a PAC or giving to candidates. And AI’s promise as a sweeping, general-purpose technology blunts Lehane’s usual playbook of crafting a core constituency around a cause like he did with the crypto voter.

But one constant remains: Lehane has no shortage of opponents, with Musk, respected California-based charities, a group of the company’s former employees, leading academics, Nobel laureates and others all lining up against the restructuring.

They worry that OpenAI’s new corporate setup will put financial interests ahead of, and divert funds away from, its charitable mission. That goes back to its 2015 founding as a nonprofit by those who believed the structure was the most responsible way to steer such powerful technology.

OpenAI originally set out last year to convert into a for-profit organization, but changed course in May amid scrutiny from state officials. The updated plan — to turn its for-profit arm into a public benefit corporation that the nonprofit controls as a stakeholder — is turning out to be no quick feat either and still requires the blessing of the attorneys general in both California and Delaware, not to mention key investor Microsoft.

Changing the structure of its for-profit arm would also eliminate the current capped-profit model.

The most pressing motivator and deadline for OpenAI is an investment led by SoftBank — $20 billion of which is contingent on the restructuring being finished by the end of the year.

State decision-makers don’t appear to be in a rush, though.

Bonta in California has hired outside help to go through OpenAI’s financials for his probe, and Jennings in Delaware plans to take similar steps. Bonta’s office met with OpenAI executives — Chief Strategy Officer Jason Kwon, head of U.S. and Canada policy Chan Park, deputy general counsel Che Chang and associate general counsel Nora Puckett.

An OpenAI spokesperson said the meeting was unrelated to the restructuring and declined to discuss specifics beyond that the company has been responsive to Bonta’s staff. A deputy attorney general for Bonta denied a records request related to the meeting, citing his ongoing investigation.

Ellie Blume, a special assistant attorney general, separately heard concerns from civic groups like the San Francisco Foundation in May and July meetings.

“They didn’t give us that much of a sense of the timeline,” said Judith Bell, the foundation’s chief impact officer, with whom Aguilar formed a coalition over the cause. The OpenAI spokesperson added the company does not have a timeline for resolution either.

Further down the line, OpenAI has ambitions to go public. Unlike nonprofits, public benefit corporations can list their shares, and OpenAI’s chief financial officer, Sarah Friar, said at the Dublin Tech Summit in May that the restructuring sets it up for a potential IPO.

Some nonprofit leaders in the state point to those financial incentives to explain OpenAI’s hiring spree.

In light of the SoftBank agreement, OpenAI could “spend a billion dollars on this and still come out $19 billion net positive,” said Tyler Johnston, the executive director of the nonprofit Midas Project, which has questioned the restructuring.

Lehane will tell you OpenAI wants to share its riches with California and grow there. Part of his mission is to lay out the stakes for the world’s fourth-largest economy if it leaves. He recently wrote directly to Newsom in a letter first reported by POLITICO, petitioning the state to change course on AI regulations or risk losing its place as the home of innovation.

OpenAI has heard pitches from state leaders across the U.S. looking to lure it away. Companies like Oracle and Musk’s Tesla, SpaceX and X Corp. have moved their headquarters to Texas while retaining a California presence.

“We want to be here,” Lehane told POLITICO. “But we also do really want to have a structure that allows us to deliver on our mission and purpose.”

The opposition

Lehane popularized the phrase “vast right-wing conspiracy” when defending the Clintons. It referred to how the internet age has allowed fringe theories to pass up to the masses, but was often used by the politicians he represented to cast themselves as the victim of a shadowy cabal.

A recurring theme of Lehane’s California operation, as many critics point out, is a persistent suspicion that rivals such as Musk are secretly behind those who have challenged OpenAI’s business plans.

OpenAI launched a countersuit against Musk’s restructuring lawsuit, alleging unfair business practices as he expands his competitor, xAI. A federal judge recently said that case could proceed.

One group the company has fixated on is the recently formed Coalition for AI Nonprofit Integrity, which backed a bill in California this year that would have effectively stopped OpenAI’s restructuring altogether. The bill was ultimately gutted and amended by its author with little explanation, but OpenAI has continued seeking to unmask the nonprofit — despite its repeated denials of any association with Musk.

O’Leary lodged a complaint with California’s top campaign finance watchdog to ask for an investigation into the group. She also subpoenaed CANI as part of the Musk lawsuit, looking for any ties. The legal filings obtained by POLITICO reveal a level of digging common in opposition research. OpenAI’s lawyers enlisted professional research services, cold-called everyone they could find with the same name as CANI’s listed leader and reviewed his property records.

Spokespeople for Musk and xAI did not respond to requests for comment.

Musk has come up during interactions between OpenAI and more established organizations in the state as well. In June, OpenAI gathered with members of Aguilar and Bell’s nonprofit coalition on a video call, ostensibly to allay concerns and explain its evolving plans.

Less than a quarter of the coalition’s 60 members joined. But it was the largest meeting between OpenAI and several of its most influential challengers in the state, the details of which POLITICO is first reporting based on the accounts of eight participants, some of whom were granted anonymity to discuss the private exchange.

The groups had asked to meet Altman, Aguilar said, and instead faced O’Leary, Moret and Mesloh. (OpenAI said in a statement that the attendees did not formally request Altman’s attendance at the meeting.)

Group leaders told POLITICO they were taken aback when the three OpenAI reps began by trying to determine if anyone had received money from Musk.

“As I’m sure you’re aware, there are lots of other organizations who are pursuing commercial interests,” an OpenAI representative said, according to contemporaneous notes taken by someone on the call and seen by POLITICO. “Some of the organizations are funded by Elon Musk.”

Two other attendees recounted that the suggestion came from O’Leary.

A “strange” and “very combative way to start,” Bell said. Catherine Bracy, CEO and co-founder of TechEquity, who was on the call, dismissed the notion that they were doing Musk’s bidding as “ludicrous and offensive.”

One participant said O’Leary was familiar with their organization from her time in Newsom’s administration, which made the company’s questioning of its background all the more puzzling.

Participants said they asked for clarification and pushed back on the notion that anyone was supported by Musk. O’Leary expressed little surprise at the groups’ denials, which two call attendees interpreted as confirmation she was already aware.

It was, a lawyer on the call would write in notes, “the first time” some OpenAI staff had witnessed the coalition’s “hostility” toward the company firsthand.

The groups left frustrated, expressing that sentiment in a letter to Bonta on June 26, telling him the OpenAI staffers could not answer the majority of their queries, and that he should keep up the investigation. The coalition argued even the revised structure that retains nonprofit control “fails to provide adequate safeguards for the public, the nonprofit’s charitable assets, and its charitable mission.”

Many later speculated that OpenAI’s portrayal of opponents as Musk-backed stooges who want to derail the company’s progress for personal gain was a ploy to undermine them and dodge tough questions about the purpose of the restructuring.

“A lot of people in California don’t like Elon Musk,” said Aguilar. “They feel like they’re taking away our credibility. It’s a poor PR tactic with really poor taste.”

Still, a few thought OpenAI’s concern was somewhat genuine, perhaps even justified, considering the bad blood between Altman and Musk.

“It's clear to me that their sense that Elon is behind this is animating a lot of their strategy,” said Bracy. “Elon’s Elon, so I guess I don't really blame them for thinking that this is something he would do.”

While the confrontations with critics are not easing opposition, OpenAI has seen some minor wins.

One of the participants on the call, the philanthropic investment firm Omidyar Network, created by eBay founder Pierre Omidyar, decided not to sign the coalition’s late June letter to Bonta, after its CEO Mike Kubzansky shared during the meeting that the group owns shares in OpenAI’s competitor Anthropic. Omidyar Network had signed all previous correspondence to Bonta.

“While our small stake in Anthropic (which was already a matter of public record) had zero bearing in our engagement with the coalition, we felt it was important to remove any doubt about the coalition’s motives,” Omidyar spokesperson Alexis Krieg said in a statement.

OpenAI declined to comment further on the meeting.

Separately, a PR firm dropped CANI as a client once it joined Actum, an influential consultancy where OpenAI adviser Butler is a partner.

The small PR firm’s co-founder, Becky Warren, had previously accused OpenAI of turning to “bullying tactics” and “doing whatever they can to distract.” CANI’s new spokesperson, Tiffany Moffatt, borrowed OpenAI’s campaign analogy to describe its strategy.

“OpenAI, which has enjoyed non-profit status under California’s tax laws, has aggressively lobbied to dismantle legislation safeguarding that status, deploying top-tier national political consultants and lawyers to intimidate and silence opposition,” she said. “This ruthless campaign has left CANI members fearful of retaliation.”

Moffatt declined to share CANI’s founders because of those fears and described the nonprofit as “a genuine grassroots coalition of concerned citizens.”

The mystery around CANI has left an opening for OpenAI to call other organizations into question. O’Leary told California’s campaign finance watchdog earlier this month that CANI may be channeling its efforts through another AI safety group in Aguilar’s coalition, Encode. Sunny Gandhi, vice president of political affairs at Encode, said her insinuation was “flat out incorrect.”

“I don’t think OpenAI truly cares that much about this group,” Aguilar said of CANI. “The message is clear: They want to paint themselves as the good guys, and everybody else, including us, we must be doing something wrong or colluding with somebody if we’re getting in their way.”

The PR plan

Yet another way OpenAI has deployed its political muscle is through a listening tour to rethink the company’s philanthropic efforts post-restructuring.

Lehane is a believer in the campaign mainstay of not just slamming the other camp’s ideas, but offering an alternative vision of his own.

“You can’t just be against everything,” he once told the Airbnb board. “You have to be for something.”

OpenAI earlier this year formed an advisory commission featuring former Newsom adviser Daniel Zingale, a close friend of O’Leary, labor icon Dolores Huerta and others from nonprofit circles to draft an independent report for its board on how the company should use its vast resources to have the largest positive impact.

In March, Mesloh and Moret previewed the panel to civic leaders by introducing Zingale at the San Francisco Foundation headquarters. The meeting was brokered by a third party, the former chair of the Latino Community Foundation, Dan Skaff, according to Aguilar.

For months, Zingale and the commission’s advisers crisscrossed the state, holding meetings to take input from different community groups. Several advocates who spoke to commission members said they perceived it as separate from the company, and Zingale said the group vigorously guarded its independence. He told POLITICO he never communicated with Altman while drafting the report.

While Zingale is not an employee, tapping him to chart a path forward for a multibillion-dollar nonprofit is one more way Lehane has sought to work the public narrative in his favor. Zingale brought on Huerta, giving the recommendations the imprimatur of the legendary co-founder of the United Farm Workers at a time when labor groups fear AI will increasingly take jobs away from workers.

Huerta, too, has decades of deep ties to Democratic politicians. She supported Harris during the 2020 primary and campaigned with her in 2024, supporting Democratic candidates dating back to Robert F. Kennedy.

She told POLITICO that she met with some of OpenAI’s researchers, but no members of the board or leadership. “We didn't ask to meet with them, and they didn't ask to meet with us,” she said.

Huerta said she had “no idea” why she was asked to sit on the commission, but guessed it was because of her civil rights background. She said she didn’t know if her agreeing to it had helped the company’s reputation in California, as it faces regulatory scrutiny. (Both of Bonta’s parents were involved in UFW organizing.)

The commission’s report strongly endorsed a nonprofit model for OpenAI with an oversight role for attorneys general and imagined the company as a new kind of philanthropic force. The advisers want OpenAI to fund a wide mix of ideas to prepare for AI’s growing presence in everyday life, from supporting public parks and other common areas as AI-free spaces to research on the technology’s safety risks.

The day of the report’s release, OpenAI held court with more than 1,000 nonprofit leaders at events across the country. It committed $50 million to fund civic group initiatives the next day, which Lehane said could come with additional AI tools, as part of his messaging on the good the company is prepared to do if regulators allow it to operate as it wants.

“Our single best lobbyist is our innovation,” Lehane said. A key piece of his approach, he added, has been to “forge partnerships with stakeholders and entities that are going to be particularly relevant to the public conversation” and show them OpenAI’s technology.

Because the report is not binding, groups that weighed in are counting on Bonta to set the course.

“This is the largest nonprofit ever,” Bell said of why there has been such intense interest in the issue. “We are talking about a scale that we simply have never seen before.”

Lehane, too, is wielding that fact publicly, as Bonta weighs what to do with his investigation and OpenAI’s billions of dollars remain, for now, headquartered in California.

OpenAI recently appealed directly to Newsom and state lawmakers, sending them a curated report about the company’s economic impact on the state, first reported by POLITICO. It was compiled by former Biden White House official Ronnie Chatterjee, now OpenAI’s chief economist, and outlined the billions in revenue the state would lose if the company packs up, reaching nearly $6 billion in total economic impact in 2030.

“There is no state or country in the world today that would not want to be California when it comes to AI,” the report stated. But it warned: “Provided, that is, that the state doesn’t drive AI out of California through well-intended but not well-understood overregulation.”