When David Sacks, President Donald Trump’s former AI and crypto czar, declared this week that public alarm over artificial intelligence is the product of a manufactured “fearmongering campaign,” he was not merely offering a soundbite — he was articulating the governing philosophy of an administration that has systematically dismantled the precautionary framework Washington spent years building around AI.
What Happened
Speaking about the trajectory of AI policy under President Trump, Sacks praised the president for “cutting through the panic” surrounding existential risks posed by advanced artificial intelligence systems. The remarks, reported by POLITICO, frame public concern — amplified by researchers, some lab executives, and bipartisan lawmakers — as an orchestrated campaign rather than a legitimate policy debate. It is a notable rhetorical escalation: not disagreement with AI safety advocates, but denial that their concerns are sincerely held.
The comments arrive at a consequential moment. Sacks, though no longer holding the formal czar title, remains one of the most influential voices in the administration’s tech orbit. POLITICO reported in May 2026 that President Trump yanked a planned AI executive order after Sacks intervened, relaying complaints from technology companies about the order’s provisions. That episode — a last-minute reversal driven by industry objections — demonstrated that Sacks’ influence did not end with his formal tenure. If anything, his role as an outside validator of Silicon Valley’s preferences has grown.
Why It Matters
The “fearmongering” framing matters because it does policy work. By casting AI risk discourse as hysteria, the administration creates political cover for a deregulatory posture that would otherwise face scrutiny. Consider the landscape: the Biden-era executive order of October 2023 invoked the Defense Production Act to require developers of powerful frontier models to report safety testing results to the government. President Trump revoked that order within days of taking office in January 2025, and the replacement framework has emphasized speed, energy infrastructure for data centers, and preemption of state-level AI regulation — not safety evaluation.
Second, the Sacks doctrine inverts the traditional relationship between industry and regulator. Historically, Washington’s anxiety has been regulatory capture — agencies going soft on the industries they oversee. Here, the dynamic is more explicit: POLITICO’s December 2025 reporting revealed that tech lobbyists themselves had grown uneasy with Sacks, fearing his aggressive tactics were “derailing” the industry’s agenda nationwide even as he championed it. When the lobbyists worry their champion is too zealous, the center of gravity in AI policy has moved decisively.
Third, dismissing existential-risk concern as a campaign ignores the unusual provenance of that concern. The scientists warning about advanced AI include figures who built the field — Geoffrey Hinton, who left Google in 2023 to speak freely about risks, and Yoshua Bengio, the most-cited computer scientist alive. The 2024 International AI Safety Report, chaired by Bengio with input from dozens of nations, documented genuine uncertainty about loss-of-control scenarios. One can dispute their conclusions; describing them as a coordinated panic campaign is a political maneuver, not a scientific one.
Historical Context
Washington has seen this playbook before. In the 1990s, tobacco industry allies spent decades reframing public-health evidence as junk science and moral panic. In the 2000s, financial industry advocates characterized warnings about derivatives as alarmism that would stifle innovation — until 2008 demonstrated otherwise. The pattern is consistent: when a powerful industry faces the prospect of precautionary regulation, the first battle is epistemological. Discredit the warning, and the regulation never arrives.
There is also a more recent parallel within tech itself. Social media companies argued for years that concerns about misinformation and teen harm were overblown moral panics; Congress never passed meaningful platform legislation, and the harms documented in internal research — most notably the Facebook Files disclosures of 2021 — went largely unaddressed. The AI debate is now following the same arc, compressed into a shorter timeframe and with far higher stated stakes.
What to Watch
Three developments will test whether the Sacks doctrine holds. First, the fate of state-level AI laws: California’s SB 1047 fight in 2024 previewed the battle, and the administration’s push for federal preemption of state AI rules will determine whether deregulation is durable or merely a Washington preference. Second, watch the frontier labs themselves — Anthropic, OpenAI, and Google DeepMind have all published safety commitments that could become liabilities if the political winds treat caution as disloyalty. Third, any incident — a model-enabled cyberattack, a bio-design misuse — would instantly stress-test the “panic” narrative, much as Deepwater Horizon ended the era of light-touch offshore drilling oversight in 2010.
Also watch Sacks’ own portfolio of influence. The May 2026 executive-order reversal showed that his channel to the Oval Office remains open. If another AI order surfaces — on preemption, procurement, or export controls — the first question in every lobbyist’s office on K Street will be the same: what does Sacks think?
Key Takeaway
The essential insight is this: the Trump administration is not simply declining to regulate AI risk — it is reclassifying the perception of risk as the problem itself. That is a far more durable form of deregulation than any executive order, because it immunizes policy against future evidence. Whether history judges that as visionary confidence or as the tobacco playbook rerun at machine speed will depend on events Washington cannot currently model — which is, of course, precisely the point the safety researchers have been making all along.
