The week ending February 28, 2026 will possible be studied in enterprise faculties for a era. Inside a single Friday afternoon, President Donald Trump banned a serious American AI firm from all federal contracts, the Pentagon designated that very same firm a nationwide safety threat on par with international adversaries, and its largest competitor swept in to seal the deal. By Sunday morning, Claude — Anthropic’s flagship AI mannequin — had leapfrogged ChatGPT on Apple’s App Retailer charts, pushed there by a wave of customers rallying behind the underdog. It was, by any measure, a unprecedented few days.
How It All Fell Aside: Anthropic and the Pentagon’s Breaking Level
The roots of the fallout stretch again months. Anthropic, the San Francisco-based AI security firm based by Dario Amodei and former members of OpenAI, had been in negotiations with the Division of Battle over a contract to deploy its Claude AI fashions on categorised navy networks. The corporate had been awarded a $200 million Pentagon contract final July — an indication, on the time, of a promising partnership.
However the negotiations had been grinding towards a wall. Anthropic’s place was clear and constant: it will assist all lawful nationwide safety makes use of of its know-how besides two — mass home surveillance of Americans, and totally autonomous deadly weapons techniques. These weren’t summary considerations. Anthropic argued that with out express contractual prohibitions, its know-how might be weaponised in ways in which no present regulation was designed to stop.
The Pentagon’s counter-position was equally agency: the navy wanted the ability to make use of AI for “all lawful functions,” full cease. Officers maintained that current regulation already prohibited mass surveillance and autonomous weapons, and that demanding further contractual carve-outs amounted to a non-public firm making an attempt to override navy command authority.
The deadline was set for five:01 p.m. on Friday, February 27. Anthropic didn’t signal.
What adopted was swift and extreme. Protection Secretary Pete Hegseth posted an announcement declaring Anthropic a “supply-chain threat to Nationwide Safety” — a designation traditionally reserved for international adversaries like Chinese language state-backed companies. Each protection contractor in the USA was now barred from doing enterprise with Anthropic. Minutes later, President Trump took to Reality Social with attribute pressure: “I’m directing EVERY Federal Company in the USA Authorities to IMMEDIATELY CEASE all use of Anthropic’s know-how. We don’t want it, we don’t need it, and won’t do enterprise with them once more!”
He wasn’t completed. In a follow-up put up, Trump warned: “Anthropic higher get their act collectively, and be useful throughout this section out interval, or I’ll use the Full Energy of the Presidency to make them comply, with main civil and legal penalties to observe.”
The language from senior Pentagon officers was no much less charged. Hegseth accused Anthropic of making an attempt to “strong-arm the USA navy into submission,” calling the corporate “sanctimonious” and “conceited.” A senior Pentagon official advised Axios that the battle had turn into private on the prime: “The issue with Dario is, with him, it’s ideological. We all know who we’re coping with.” Trump was much more blunt in a Reality Social tirade, calling Anthropic an “out-of-control, Radical Left AI firm run by individuals who don’t know what the true World is all about.”
Anthropic Fights Again: “We Are Patriotic Individuals”
Anthropic’s response was defiant however measured. In an unique interview with CBS Information, CEO Dario Amodei pushed again laborious in opposition to the characterisation of his firm as ideologically pushed or obstructionist.
“We’re patriotic Individuals,” Amodei mentioned. “All the things we now have executed has been for the sake of this nation, for the sake of supporting US nationwide safety. Our leaning ahead in deploying our fashions with the navy was executed as a result of we consider on this nation.”
Amodei known as the federal government’s actions “retaliatory and punitive,” and Anthropic issued a proper assertion pledging to problem the supply-chain threat designation in court docket. The corporate known as the transfer “legally unsound” and warned it will set a “harmful precedent for any American firm that negotiates with the federal government.”
Anthropic additionally disputed the declare that it had acquired honest discover of the ultimate phrases. “We’ve not but acquired direct communication from the Division of Battle or the White Home on the standing of our negotiations,” the corporate mentioned, including that the brand new contract language proposed by the Pentagon would enable its security guardrails to be “disregarded at will.”
Earlier that Friday, forward of the deadline, senior members of the Senate Armed Companies Committee had privately urged each Hegseth and Amodei to increase negotiations and work with Congress to discover a decision. That plea went unanswered.
OpenAI Strikes In — and Instantly Attracts Hearth
Inside hours of the ban on Anthropic, OpenAI CEO Sam Altman posted on X that his firm had “reached an settlement with the Division of Battle to deploy our fashions of their categorised community.” The announcement was swift, the timing was inconceivable to disregard, and the backlash was instant.
The central query that critics requested — loudly — was this: if OpenAI maintained the identical security crimson strains as Anthropic, how had it managed to achieve an settlement the identical day Anthropic was ejected? Altman’s personal weblog put up acknowledged that OpenAI’s deal prohibited home mass surveillance, autonomous weapons techniques, and “high-stakes automated selections.” These positions sounded remarkably much like what Anthropic had been insisting upon for months.
In an announcement OpenAi mentioned, “Yesterday we reached an settlement with the Division of Battle for deploying superior AI techniques in categorised environments, which we requested they make out there to all AI corporations” supply: X
Altman, to his credit score, didn’t cover from the awkwardness. In an “ask me something” session on X over the weekend, he acknowledged the optics bluntly: the deal had been “positively rushed,” and he admitted “the optics don’t look good.”
So why do it? “We actually wished to de-escalate issues, and we thought the deal on supply was good,” Altman wrote. “If we’re proper and this does result in a de-escalation between the DoW and the business, we’ll appear to be geniuses, and an organization that took on loads of ache to do issues to assist the business. If not, we’ll proceed to be characterised as rushed and uncareful.”
In a memo to his personal crew — the substance of which grew to become public — Altman framed it as a matter of precept over optics: “It is a case the place it’s vital to me that we do the proper factor, not the straightforward factor that appears sturdy however is disingenuous. However I notice it could not ‘look good’ for us within the brief time period, and that there’s a lot of nuance and context.”
Critics weren’t solely satisfied. Tech publication Techdirt argued that OpenAI’s contract language, which ties knowledge assortment to Government Order 12333, “completely does enable for home surveillance” as a result of the order governs how the NSA legally collects communications — together with these of Americans — by intercepting knowledge outdoors US borders. OpenAI disputed the characterisation, with a spokesperson stating: “Publicly out there info can solely be utilized by the navy for protection and intelligence functions if it’s tied to authorised national-security missions.”
Altman additionally took the weird step of defending Anthropic immediately, calling the supply-chain threat designation an “extraordinarily scary precedent” for any American firm that negotiates in good religion with the federal government.
The Fallout: Market Indicators and an Business on Edge
The market didn’t look forward to authorized arguments to play out. By Saturday, Anthropic’s Claude had overtaken OpenAI’s ChatGPT in Apple’s App Retailer rankings — a hanging client sign that the general public’s sympathies could lie with the corporate that held the road, not the one which closed the deal.
Claude has overtaken OpenAI’s ChatGPT in Apple’s App Retailer rankings, Supply: X
Throughout the AI business, workers at each OpenAI and Google had been reportedly calling on their management to talk out in opposition to the Anthropic ban. The episode has opened a broader debate about how AI corporations ought to interact with authorities contracts, and what occurs when security rules collide with political stress.
The core stress — between a authorities that wishes AI instruments with out circumstances and corporations that consider some circumstances are non-negotiable — is unlikely to be resolved by any single contract. As Altman himself noticed on X: “I feel Anthropic could have wished extra operational management than we did. We and the DoW bought comfy with the contractual language, however I can perceive different folks would have a distinct opinion right here.”
Anthropic, for its half, is heading to court docket. The corporate has vowed to problem the supply-chain threat designation, calling Hegseth’s authority to increase the ban past Pentagon contractors legally questionable. It has framed the battle not simply as a matter of company curiosity, however of constitutional precept: “We consider this designation would each be legally unsound and set a harmful precedent for any American firm that negotiates with the federal government.”
What is evident is that the foundations of engagement between Silicon Valley and Washington have shifted, probably completely. The query of who will get to determine how AI is utilized in battle — a know-how firm’s ethics board, or a navy chain of command — is now not hypothetical. It’s now a reside authorized and political battle, taking part in out in actual time.
For Anthropic, the following transfer is the courtroom. For OpenAI, it’s proving that its rushed deal holds as much as scrutiny. And for the remainder of the AI business, the lesson could also be easier: in Washington proper now, the politics of AI are transferring sooner than anybody’s security pointers.
Jason Jones Jason Jones Read More
