Anthropic, the artificial intelligence company behind the Claude chatbot, has pushed for greater government regulation of AI in ways that echo elements of the Washington playbook employed by convicted fraudster Sam Bankman-Fried during his time as a cryptocurrency entrepreneur, a review of public records shows.
asserted
review → push → records
Both Anthropic and Bankman-Fried poured millions of dollars into Washington influence campaigns while publicly arguing that government rules were needed to make their respective industries safer, although the money flowed through notably different channels.
asserted
money → pour → channels
Critics in both industries have accused them of using regulation to strengthen their competitive positions.
asserted
Critics → accuse → positions
Bankman-Fried’s political operation later became part of a criminal case involving illegal campaign contributions, while Anthropic’s spending has consisted of disclosed lobbying and advocacy efforts.
asserted
spending → become → efforts
Anthropic, for its part, denies this characterization.
asserted
Anthropic → deny → characterization
Bankman-Fried, meanwhile, later said his public posture on regulation was essentially "just PR" in a social media direct-message exchange published days after FTX collapsed.
asserted
FTX → say → exchange
Anthropic has been pushing Washington to put guardrails around the fast-moving AI industry, warning that increasingly powerful models could pose serious national security and public safety risks.
uncertain
models → push → risks
CEO Dario Amodei has called for mandatory testing and auditing of advanced models, while the company has spent millions lobbying policymakers and backing groups that support tougher AI rules.
asserted
that → call → rules
The push comes as Anthropic is increasingly engaging with a Trump administration that has made rapid AI development and U.S. dominance in the technology a major priority, warning that falling behind could carry economic and national security consequences.
uncertain
falling → come → consequences
Amodei joined other major tech executives at the White House on Sept. 29, where they signed a voluntary agreement outlining safety standards for companies developing powerful AI systems.
asserted
they → join → systems
Before being convicted of fraud, Bankman-Fried lobbied Washington for a regulatory system for crypto led by the Commodity Futures Trading Commission, presenting it as necessary to protect customers and bring clarity to the industry.
asserted
Fried → convict → industry
He was a prominent supporter of the Digital Commodities Consumer Protection Act, legislation that would have created such a framework — with campaign finance records showing that he cut checks to the bill’s sponsors in the Senate.
asserted
he → create → Senate
At the time, some crypto advocates opposed the DCCPA as overly favorable to centralized exchanges such as FTX.
asserted
advocates → oppose → FTX
Prosecutors later alleged that Bankman-Fried and his co-conspirators flooded the political system with tens of millions of dollars in illegal contributions as part of an effort to increase FTX’s influence in Washington and help advance favorable legislation.
asserted
Fried → allege → legislation
Bankman-Fried justified his efforts on the grounds that public safety was at stake, using similar rhetoric to that employed by Anthropic in its quest to regulate AI development.
"We need regulatory oversight and customer protection," he said on social media weeks before FTX collapsed.
asserted
FTX → justify → media
Testifying before Congress, he argued that crypto regulation should provide "system safety and soundness."
asserted
regulation → testify → safety
Anthropic has also pursued a lobbying strategy that calls for government regulation, warning lawmakers that advanced AI poses serious national security and public safety risks.
asserted
AI → pursue → risks
Amodei told the Senate Judiciary Committee in 2023 that AI poses "extraordinarily grave threats to US national security" and called for a mandatory "testing and auditing regime" with the power to block deployment of models found unsafe.
"
asserted
AI → tell → models
Do you guys remember what Sam Bankman-Fried was doing when he was caught red-handed?
asserted
he → remember → what
Aggressively lobbying Washington to ban his competitors," Brian Chau, the founder of a news platform that leverages AI to aid in investigations, said in response to a piece written by Amodei calling for federal AI regulation.
asserted
that → lobby → regulation
"The apple doesn't fall far from the tree."
asserted
apple → fall → tree
Anthropic's pro-safety rhetoric, much like Bankman-Fried’s, is backed by large amounts of cash.
asserted
rhetoric → back → cash
The firm has spent close to $7 million on lobbying efforts from 2025 through the middle of 2026, per federal disclosures, and has set aside $40 million to Public First Action, a 501(c)(4) tied to the super PAC Public First, which is being spent to advance pro-regulation talking points.
asserted
which → spend → points
"Anthropic is running a sophisticated regulatory capture strategy based on fear-mongering.
asserted
Anthropic → run → mongering
It is principally responsible for the regulatory frenzy that is damaging the startup ecosystem," former Trump administration AI advisor David Sacks posted to X in October 2025, responding to Anthropic’s strategy.
asserted
Sacks → damage → strategy
Much like some crypto advocates argued that the Bankman-Fried-backed DCCPA would have harmed his smaller competitors, Anthropic has been accused of supporting bills, namely California’s SB 53, that allegedly would harm less established AI labs.
uncertain
that → argue → labs
In the face of criticism, Anthropic has maintained that its commitment to public safety is genuine and pushed back against claims that its legislative activity is disadvantageous to smaller competitors.
"Some have suggested that we are somehow interested in harming the startup ecosystem.
uncertain
we → maintain → ecosystem
Startups are among our most important customers," Amodei wrote in October 2025, noting that SB 53 exempted companies with gross revenues below $500 million.
asserted
SB → write → million
While labs with revenue below the threshold are partially exempted, lower-income firms that use a certain amount of computational resources are still subject to transparency requirements under SB 53.
asserted
that → exempt → SB
Brian Chau, founder and CEO of the AI-assisted investigative publication Effort News, has also compared Anthropic’s public messaging about unreleased models to Bankman-Fried’s messaging as FTX unraveled.
asserted
FTX → assist → messaging
"I remember when FTX was going down SBF would make up a new cope narrative every week and all the FTX stans would instantly believe him," Chau wrote in response to an article about an unreleased Anthropic model.
asserted
Chau → remember → model
Anthropic and representatives for Bankman-Fried did not respond to multiple requests for comment when reached by Fox News Digital.
asserted
Anthropic → respond → Digital
The Aporia
