Defense Secretary Pete Hegseth summoned Anthropic CEO Dario Amodei to the Pentagon on Tuesday and delivered an ultimatum: lift the restrictions on Claude's military use by Friday evening, or face termination of a $200 million contract, designation as a supply-chain risk, or invocation of the Defense Production Act to seize access to the technology anyway.

The trigger? Pentagon officials believe Anthropic — through its work with Palantir — inquired whether Claude had been used in a January operation that captured Venezuelan leader Nicolás Maduro. The Pentagon interpreted that as Anthropic attempting to police how a sovereign government uses its AI.

Amodei didn't flinch. He reiterated Anthropic's red lines: no autonomous weapons, no mass surveillance. Sources say the company has no plans to budge.

This is the first time the U.S. government has directly threatened an AI company with national-security-level consequences over safety guardrails. The Defense Production Act — a Korean War-era law — would let the Pentagon compel access to Claude regardless of Anthropic's terms of service.

This isn't a policy debate. It's a precedent. If the government can force an AI company to remove safety restrictions under threat of blacklisting, every AI lab's guardrails become negotiable. And if Anthropic holds the line, it proves that an AI company can say no to the most powerful buyer on Earth. Either outcome reshapes the industry. Watch Friday.

• Meta Signs Up to $100B AMD Chip Deal — Days After Its Nvidia Deal: Meta struck a multiyear agreement with AMD for 6 gigawatts of AI chips — worth up to $100 billion. AMD issued Meta a warrant to buy 160 million shares. Shipments of AMD's MI450 GPUs start in H2 2026. This comes one week after Meta's massive Nvidia deal. Action: Meta is forcing chip competition. If you're building on AI, inference costs are heading down — plan your infrastructure spend accordingly.

• Profound Hits $1B Unicorn Helping Brands Control What AI Says About Them: Profound raised $96M at a $1B valuation. It monitors how ChatGPT, Claude, Gemini, and Perplexity represent your brand in AI-generated answers. Already serves 10% of the Fortune 500 including Target, Walmart, and Figma. Action: Search your brand name in ChatGPT and Claude right now. What comes back is what your customers are seeing — and you're probably not managing it yet.

• Samsung Unpacked Today: The First Phone With Three AI Agents Built In: Samsung launches the Galaxy S26 today with Perplexity ("Hey Plex"), Google Gemini, and an overhauled Bixby — three AI agents embedded into one device. The S26 lets you switch agents based on the task without leaving the OS. Action: Whether you buy one or not, this is the new standard. If your product or content isn't optimized for AI assistants on phones, you're missing where discovery is heading.

The "Friday Deadline" Drill.

Open ChatGPT or Claude. Paste this:

"Pretend I have until Friday to make one decision I've been postponing. I'll tell you what it is. Your job: give me the exact first step I need to take in the next 60 minutes to make it irreversible. No planning. No research. Just the action that makes turning back harder than moving forward."

Then name the decision. The best deadlines aren't the ones someone gives you. They're the ones you set — and refuse to negotiate.

Break Your Limits. Build Your Legacy.

The Limitless Insider — Daily Edition

www.islimitless.com