A public fight over AI safety has moved beyond policy disagreement and into accusations of hidden motives, legal pressure, and retaliation fears. The dispute centers on whether safety-focused groups are raising legitimate concerns about artificial intelligence or helping drive regulation that could benefit some companies while burdening others.
The latest flashpoints involve White House AI and Crypto Czar David Sacks, OpenAI chief strategy officer Jason Kwon, Anthropic, and nonprofits that advocate for responsible AI policy. Together, the episodes show how tense the debate has become as AI companies, policymakers, and critics argue over the future shape of AI regulation.
Why AI safety groups feel targeted
Several AI safety groups told TechCrunch that recent comments from Sacks and OpenAI fit a broader pattern of Silicon Valley trying to pressure its critics. The groups argue that this is not the first time the industry has pushed back aggressively against safety proposals.
In 2024, some venture capital firms spread rumors that California’s AI safety bill, SB 1047, would send startup founders to jail. The Brookings Institution later described that claim as one of many “misrepresentations” about the bill. Governor Gavin Newsom ultimately vetoed it anyway.
The current dispute has created enough concern that many nonprofit leaders contacted by TechCrunch asked to speak anonymously to avoid retaliation against their organizations. That fear is itself part of the story: groups that work on AI safety say the pressure is changing how openly they can criticize powerful companies.
David Sacks takes aim at Anthropic
On Tuesday, Sacks wrote a post on X accusing Anthropic of using fear about AI to advance regulation that would help itself and burden smaller companies. Anthropic has raised concerns about AI’s ability to contribute to unemployment, cyberattacks, and catastrophic harms to society.
Sacks was responding to a viral essay from Anthropic co-founder Jack Clark about his fears regarding AI. Clark had delivered the essay as a speech at the Curve AI safety conference in Berkeley weeks earlier.
Anthropic was the only major AI lab to endorse California’s Senate Bill 53, or SB 53. The bill sets safety reporting requirements for large AI companies and was signed into law last month.
Anthropic is running a sophisticated regulatory capture strategy based on fear-mongering. It is principally responsible for the state regulatory frenzy that is damaging the startup ecosystem.
Sacks also said in a follow-up post on X that Anthropic has positioned “itself consistently as a foe of the Trump administration.” The accusation suggests that, in his view, Anthropic’s public safety concerns are not just technical warnings but part of a broader political and regulatory strategy.
OpenAI defends subpoenas to nonprofits
OpenAI is facing its own controversy over how it is responding to outside criticism. Jason Kwon wrote a post on X explaining why OpenAI was sending subpoenas to AI safety nonprofits, including Encode, a nonprofit that advocates for responsible AI policy.
Kwon connected the subpoenas to Elon Musk’s lawsuit against OpenAI. Musk has argued that the ChatGPT maker has moved away from its nonprofit mission. Encode filed an amicus brief in support of Musk’s lawsuit, while other nonprofits publicly criticized OpenAI’s restructuring.
“This raised transparency questions about who was funding them and whether there was any coordination,” said Kwon.
NBC News reported this week that OpenAI sent broad subpoenas to Encode and six other nonprofits that criticized the company. The subpoenas sought communications related to two of OpenAI’s biggest opponents, Musk and Meta CEO Mark Zuckerberg. OpenAI also asked Encode for communications connected to its support of SB 53.
Inside OpenAI, the episode appears to have created unease. One prominent AI safety leader told TechCrunch that there is a widening divide between OpenAI’s government affairs team and its research organization. OpenAI’s safety researchers frequently publish reports about AI risks, while the company’s policy unit lobbied against SB 53 and argued for uniform federal rules instead.
OpenAI’s head of mission alignment, Joshua Achiam, also criticized the subpoenas in a post on X.
“At what is possibly a risk to my whole career I will say: this doesn’t seem great,” said Achiam.
The conspiracy claim and the response
Brendan Steinhauser, CEO of the AI safety nonprofit Alliance for Secure AI, told TechCrunch that OpenAI seems to believe its critics are part of a Musk-led conspiracy. His organization has not been subpoenaed by OpenAI.
Steinhauser rejected that idea and said much of the AI safety community is also critical of xAI’s safety practices, or lack thereof. In his view, the pressure is meant to discourage nonprofits from speaking out.
“On OpenAI’s part, this is meant to silence critics, to intimidate them, and to dissuade other nonprofits from doing the same,” said Steinhauser. “For Sacks, I think he’s concerned that [the AI safety] movement is growing and people want to hold these companies accountable.”
Sriram Krishnan, the White House’s senior policy advisor for AI and a former a16z general partner, also joined the debate with a social media post. He called AI safety advocates out of touch and urged them to speak with “people in the real world using, selling, adopting AI in their homes and organizations.”
What the debate says about AI regulation
The conflict highlights a hard trade-off for the AI industry. Companies want to build AI into large consumer products, while safety advocates want stronger accountability before harms become harder to manage.
Public concern is not limited to the catastrophic risks often emphasized by the AI safety movement. A recent Pew study found that roughly half of Americans are more concerned than excited about AI, though it is unclear exactly what worries them. Another recent study found that American voters care more about job losses and deepfakes than catastrophic risks caused by AI.
That difference matters because the safety debate is not only about whether AI should be regulated. It is also about which risks deserve the most attention, who gets to define them, and whether rules should be set by states or handled uniformly at the federal level.
For Silicon Valley, the worry is that safety rules could slow rapid growth. With AI investment supporting much of America’s economy, fear of overregulation is understandable. But after years of largely unregulated AI progress, the AI safety movement appears to be gaining momentum heading into 2026.
The backlash from powerful industry figures may therefore signal more than frustration. It may show that safety-focused groups are becoming harder for the AI industry to ignore.