⛔ A new chapter in the AI war: In an escalation between Silicon Valley and the Pentagon, Anthropic filed a lawsuit against the US Department of Defense (DoD) on Monday after the Trump administration blacklisted the AI firm as a “national security threat.” This is the culmination of a week-long standoff between the Trump administration and the creators of Claude over the military applications of GenAI. So, what happened exactly?

The genesis

When Anthropic built Claude, it baked in strict guardrails against autonomous weaponry and mass surveillance. These ethical red lines have now collided head-on with the Pentagon’s new directive. In late February, the DoD demanded the right to deploy AI for all “lawful” military purposes — including advanced combat applications.

Defense Secretary Pete Hegseth issued an ultimatum: comply… or face the consequences. Hegseth explicitly threatened to invoke extraordinary executive powers to seize the company’s proprietary tech — all in the name of national security, naturally.

Despite Anthropic’s previous USD 200 mn agreement to integrate its tech into classified networks last June, CEO Dario Amodei has refused to budge on the company’s safety protocols. Tensions reportedly peaked following the January operation in Venezuela involving the abduction of Nicolas Maduro, where US forces relied heavily on AI-backed logistics.

Trump has since ordered all federal agencies to terminate their use of Claude, branding Anthropic a leftist organization. Ironically, reports from both The Wall Street Journal and Axios suggest the military may have already utilized Claude in recent operations involving Iran — particularly for intelligence analysis, target selection, and field simulations.

The tech divide

OpenAI saw a chance… and took it. The Microsoft-backed rival quickly signaled its cooperation with the Pentagon, securing a specialized agreement to deploy its models across the DoD’s classified infrastructure. CEO Sam Altman defended the move, stating that the Pentagon’s framework aligns with OpenAI’s principles regarding human accountability in weapons systems.

The backlash: This triggered a massive PR crisis for OpenAI. The QuitGPT movement — already fueled by news that co-founder Greg Brockman donated USD 25 mn to MAGA — has gained endorsements from high-profile figures, such as celebrated actor Mark Ruffalo.

Public sentiment is shifting rapidly: Early data shows nearly 2 mn users have ditched ChatGPT for Claude, citing a preference for the latter’s ethical stance. It’s horrible timing for OpenAI, too, which is reportedly on track to lose USD 14 bn in 2026 as it loses market share to its rivals.

Trouble ahead

Government defense contracts represent a stream of bns in sustainable revenue that tech giants simply cannot ignore. These agreements, however, are a double-edged sword that many investment funds now view with trepidation. For these investors, the risk is clear: any association with controversial military applications can shatter reputations, leading to massive losses in market share as socially conscious consumers and partners jump ship.

Beyond the immediate financial damages — estimated to be in the mns — the Anthropic lawsuit argues that the government’s blacklist is an arbitrary interference in corporate governance, designed to stifle competition and scare off investors. This isn’t just about one company; it’s a threat to the entire AI ecosystem. Within hours of the filing, dozens of employees from Google DeepMind — and even OpenAI — voiced their support for Anthropic.

A canary in the coal mine for Big Tech: Giants like Microsoft, Google, Meta, and Amazon are now forced to re-evaluate their military ties to avoid the same wrath OpenAI is facing. As the Trump administration prepares to potentially weaponize the Defense Production Act to force compliance, this case will set quite the precedent, determining the future trajectory of the relationship between US sovereignty and the independence of AI firms.

(** Tap or click the headline above to read this story with all of the links to our background as well as external sources.)