OpenAI accused of using intimidation tactics to erode California’s proposed AI safety law: Here’s what we know

A small nonprofit organization has accused OpenAI, led by Sam Altman, of using scare tactics to water down California’s AI security law, known simply as SB 53.

The accusation came in a viral post on social media platform X (formerly Twitter) by 29-year-old Nathan Calvin, Encode’s legal counsel. Calvin claimed that OpenAI was working to undermine the California Border AI Transparency Act (SB 53), which is being debated by lawmakers.

In his own lengthy post addressing the issue, OpenAI’s chief strategy officer Jason Kwon denied the allegations made by Calvin and noted that Encode’s funding was questioned, implying that the public accusation was funded by Elon Musk.

What are the allegations?

Calvin claimed that OpenAI was using its legal battle against Elon Musk to intimidate critics, including Encode. He also clarified in his post that Encode did not receive funding from Elon Musk.

“I believe OpenAI is using the pretext of its lawsuit against Elon Musk to scare critics and imply that Elon is behind it all,” he said.

According to the Fortune report, OpenAI claimed to have written a letter to California Governor Gavin Newsom’s office in an attempt to ease the SB 53 requirement that mandates companies to publish a public edge AI framework and a transparency report when deploying or significantly modifying a model.

He said he is speaking now because SB 53 negotiations are still ongoing and wants to wait until the bill is signed. Another reason he said was because OpenAI’s Chris Lehane claimed in a LinkedIn post that the company was “working to improve” SB 53.

Calvin called the experience “the most stressful time of my professional life” and noted that OpenAI’s products and research “deserve real praise,” adding that many of the company’s employees care about being a force for good. “I want to see this aspect of OAI, but instead I see them trying to intimidate and silence critics. Does anyone believe these actions are consistent with OpenAI’s not-for-profit mission of ensuring that AGI benefits humanity?” he asked.

Concerns expressed about allegations

Calvin’s accusations, in particular, drew backlash from current and past employees of OpenAI, including head of mission orchestration Joshua Achiam, who called the developments “not very good” in a personal thread on X.

He shared Calvin’s post: “While it’s a risk to my entire career, I will say this: it doesn’t look good. Lately I’ve been describing my role as something along the lines of “public defender,” so I’d be remiss if I didn’t share some thoughts for the public on this subject. Some thoughts on the subject…”

In his post, Achiam called on OpenAI to distinguish between the powerful Elon Musk and constructive critics. “We cannot be doing things that will turn us into a frightening force instead of a virtuous force. We have a duty and a mission towards all humanity, and the bar for fulfilling this duty is quite high,” he said.

He wasn’t alone. Former OpenAI board member Helen Toner, who resigned after the unsuccessful 2023 ouster of Sam Altman as chairman, said that while the company was doing great things, “the dishonesty and intimidation tactics in its policy work are not what it is.”

Tyler Johnston, founder of AI watchdog group Midas Project, also shared an experience of alleged intimidation in a response to Calvin’s post, saying: “[I] There was a knock on my door in Oklahoma with every text/email/request for documents regarding OpenAI’s management and investors, ‘in the broadest sense permitted’.”

Johnston noted that the goal appeared to be to check whether Elon Musk was funded, adding: “If they had just asked if I was funded by Musk, I would have been happy to give them a simple ‘man if only’ and call it a day. Instead, practically speaking, I would have compiled a list of every journalist, congressional office, partner organization, former employee, and member of the public we talked to about their restructuring.” they wanted.”

How did OpenAI react?

Neither Sam Altman nor OpenAI’s official channels have commented yet. Kwon approached X to question the motivation behind the allegations.

“We wanted to know, and we still wonder, whether Encode was working in collaboration with third parties who had commercial competitive interests adverse to OpenAI. The stated narrative makes it sound as if it is something it is not.” Kwon said. He also included an excerpt from the subpoena, which he said shows all requests for documents made by OpenAI.

In September, the San Francisco Standard reported that OpenAI had subpoenaed its most vocal critics in August over concerns that Elon Musk was funding negative reviews to target the company.

Elon Musk and OpenAI, in particular, are engaged in an ugly, public battle over the company’s governance and its shift towards for-profit structure.