La corte d'appello degli Stati Uniti conferma il divieto del Pentagono sul modello Claude di Anthropic
A federal appeals court affirmed the Pentagon’s decision to bar Anthropic from its supply chain after the startup refused to lift safeguards on its Claude AI system.
The AFP report adds new details to the ongoing story, including the court’s reference to an Anthropic executive’s objection to Palantir’s use of Claude during the operation that captured Venezuelan President Nicolás Maduro, direct quotes from Under Secretary of Defense Emil Michael, and the contrast with a separate California federal court decision that found a broader Trump‑era ban illegal. These specifics deepen understanding of the legal rationale and the broader policy conflict surrounding AI safety safeguards.
The appeals court affirmed the Pentagon’s supply‑chain risk designation against Anthropic, keeping the company barred from military contracts after it refused to lift safeguards on its Claude model.
The appeals court affirmed the Pentagon’s ban on Anthropic, citing the company’s refusal to lift safety restrictions on Claude and an internal objection to its use by Palantir during the Maduro capture operation, thereby solidifying the government’s authority to blacklist AI vendors for national‑security reasons.
The AFP report adds new context to the court’s decision, citing the Anthropic executive’s objection to Palantir’s use of Claude during the Maduro capture operation, quoting Under Secretary Emil Michael’s safety concerns, and noting the clash with a separate California court ruling. It also references the broader industry reaction, including other AI firms securing Pentagon contracts and the political backdrop of White House criticism of Anthropic’s CEO.
Cosa è successo
The U.S. Court of Appeals for the Federal Circuit ruled Friday that the Pentagon can keep Anthropic on a national‑security blacklist, maintaining the ban on the company’s Claude model for military use.
The appellate panel, led by Judge Gregory Katsas, affirmed the Pentagon’s March designation that labeled Anthropic a supply‑chain risk after the company refused Secretary Pete Hegseth’s request to remove restrictions on Claude’s use for fully autonomous lethal weapons and mass domestic surveillance.
The decision notes that an Anthropic executive raised concerns about the model’s use by a Pentagon contractor, Palantir, during the operation that captured Venezuelan President Nicolás Maduro on Jan. 3, which the Under Secretary of Defense Emil Michael said “raised material doubts” about warfighter safety.
The court held that the President and Secretary of Defense have the authority to balance such risks, and that Anthropic cannot retroactively modify models already delivered to the military, even though the company controls future version behavior.
The ruling contrasts with a separate California federal court decision that found a broader Trump‑era ban on Anthropic illegal, highlighting a split in judicial interpretation of the administration’s AI restrictions.
The ruling clarifies the government’s authority to label AI providers as supply‑chain risks, setting a precedent for future restrictions on AI technologies deemed unsafe for defense. It also underscores tensions between AI firms and U.S. security agencies over model safeguards, and may affect Anthropic’s upcoming IPO and its standing with other federal contractors.
By upholding the blacklist, the court reinforces the Pentagon’s ability to exclude AI vendors that do not comply with specific security or ethical constraints, potentially prompting other contractors to demand similar safeguards from AI providers.
The decision may influence the broader regulatory landscape for AI in defense, encouraging the development of formal guidelines for model safety, export controls, and the handling of autonomous weapon systems.
Anthropic’s status as a high‑valued AI startup—valued near $1 trillion and poised for a blockbuster IPO—means the ruling could affect investor confidence and the company’s market positioning, especially as rivals like OpenAI, xAI, Google, and Microsoft have secured Pentagon agreements.
Interactive Mechanism
Meccanismo interattivo: come funziona realmente
Esplora la tecnologia alla base di questo sviluppo in modo interattivo.
Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
Guardrail & Verification:🛡️ Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
Verifica concettuale interattiva+10 Points
AI Ethics Quiz
Impossibility results in algorithmic fairness (e.g. Kleinberg et al., Chouldechova) show what?
Cosa guardare dopo
Future litigation over the conflicting California federal court decision, Pentagon procurement policy revisions, Anthropic’s IPO timeline, and how other AI firms respond to the precedent.
Legal: Anthropic has indicated it may appeal and is reviewing “all options, including further review,” suggesting possible higher‑court challenges that could reshape the scope of the blacklist.
Policy: The Pentagon may issue updated procurement rules clarifying the criteria for labeling AI firms as supply‑chain risks, which could affect future contracts across the defense sector.
Market: Analysts will monitor Anthropic’s IPO timeline and pricing, as the blacklist could impact valuation and demand from institutional investors wary of regulatory risk.
Industry response: Other AI firms may pre‑emptively adjust model safeguards or negotiate separate agreements with the Department of Defense to avoid similar restrictions.
Questa storia canonica viene aggiornata quando l'evento in via di sviluppo cambia materialmente. Il suo URL e la data di pubblicazione originale non cambiano mai.
The AFP report adds new context to the court’s decision, citing the Anthropic executive’s objection to Palantir’s use of Claude during the Maduro capture operation, quoting Under Secretary Emil Michael’s safety concerns, and noting the clash with a separate California court ruling. It also references the broader industry reaction, including other AI firms securing Pentagon contracts and the political backdrop of White House criticism of Anthropic’s CEO.
The appeals court affirmed the Pentagon’s ban on Anthropic, citing the company’s refusal to lift safety restrictions on Claude and an internal objection to its use by Palantir during the Maduro capture operation, thereby solidifying the government’s authority to blacklist AI vendors for national‑security reasons.
The appeals court affirmed the Pentagon’s supply‑chain risk designation against Anthropic, keeping the company barred from military contracts after it refused to lift safeguards on its Claude model.
The AFP report adds new details to the ongoing story, including the court’s reference to an Anthropic executive’s objection to Palantir’s use of Claude during the operation that captured Venezuelan President Nicolás Maduro, direct quotes from Under Secretary of Defense Emil Michael, and the contrast with a separate California federal court decision that found a broader Trump‑era ban illegal. These specifics deepen understanding of the legal rationale and the broader policy conflict surrounding AI safety safeguards.