Pentagon Drops Anthropic, Apple Ends ChatGPT Monopoly, and the EU AI Act Is 55 Days Away

Every week, AI rewrites the rules. This is your weekly briefing. 🧠

The Pentagon just fired its artificial intelligence provider. Because it was too safe.

This week delivered three developments that every business leader in Europe needs to understand - because each one changes the rules of the game your company is playing right now.

The Pentagon Replaced Anthropic With OpenAI Over Safety Guardrails

The United States Department of Defense officially replaced Anthropic’s Claude across its classified military systems this week. The reason was not performance. It was not cost. It was safety.

Anthropic refused to remove the guardrails that block autonomous weapons systems and mass surveillance applications. The Department of Defense responded by labelling those guardrails a supply chain risk to national security. Within hours, OpenAI signed a new contract to fill that gap.

This is one of the most significant moments in the short history of commercial AI. A major government just established on the record that AI safety features are a liability - not an asset - when national security priorities are involved. The implications for how AI gets regulated, deployed, and constrained globally will ripple for years.

Apple Ends ChatGPT’s Two-Year Monopoly on iPhone

Apple opened its Worldwide Developers Conference today with the biggest shift in iPhone history. iOS 27 breaks open the personal AI market on Apple devices.

From this autumn, over two and a half billion Apple devices worldwide will give users the choice between ChatGPT, Gemini, and Claude as their built-in personal artificial intelligence. Every iPhone becomes a direct battleground between the three largest AI companies on the planet.

For businesses building products, services, or customer experiences on Apple’s ecosystem, this changes the assumptions you were working with. The user is no longer defaulting to one AI. They are choosing. And the experience your product delivers will be compared against all three.

The EU AI Act Enforcement Deadline Is Now 55 Days Away

Underneath all of this, a deadline is approaching that many European companies are not ready for. The European Union AI Act enforcement deadline is 55 days away.

Companies operating in Europe that are not compliant are running out of time. This is no longer a future obligation - it is an immediate operational risk. Non-compliance carries significant penalties, and the enforcement mechanisms are already in place.

If your business uses AI in customer interactions, hiring, credit decisions, content moderation, or any high-risk category as defined by the Act, the clock is running.

What the next 90 days mean for your business

These three developments are not separate stories. They are part of the same shift.

AI is no longer something you explore or experiment with at the margins of your business. The decisions made in the next 90 days - on compliance, on infrastructure, on which AI partners you work with and why - will define where your business stands for the next five years.

The companies that treat this moment as urgent will have a structural advantage over those that treat it as background noise.

At Aliando, we help businesses navigate exactly this kind of inflection point turning complexity into a clear strategy before the window to move first closes.

Follow for your weekly AI intelligence.

(Do not miss the video below, where each of these developments is broken down in under a minute.)