One AI CEO Said No to the Pentagon


One AI CEO Said No to the Pentagon

This week, the biggest story in AI isn't a new tool — it's a fight that could change which tools you're allowed to use.


The Big Story

Anthropic, the company behind Claude, just got blacklisted by the Pentagon. The reason: CEO Dario Amodei refused to let the military use Claude for autonomous weapons and mass domestic surveillance without human oversight. Defense Secretary Pete Hegseth gave Amodei a Friday 5 p.m. deadline to comply. He didn't. Within hours, Trump signed an order directing all federal agencies to stop using Anthropic products, and the Pentagon designated the company a "supply chain risk" — a legal label normally reserved for foreign adversaries like Huawei.

Here's the twist that matters: OpenAI struck a Pentagon deal the same day with the exact same safety restrictions Anthropic was blacklisted for. Sam Altman confirmed OpenAI's deal includes prohibitions on autonomous weapons and mass surveillance — the same two red lines Amodei drew. The Pentagon accepted those terms from OpenAI while calling them "fundamentally incompatible with American principles" from Anthropic.

What this means for you: If your business has any federal contracts, you may need to audit your AI stack. The Pentagon's order bars contractors from using Claude even in unrelated workflows. More broadly, this is the first time a U.S. AI company has been treated like a Chinese telecom firm — not for espionage, but for maintaining safety policies. Your AI vendor choice just became a political risk factor, and that's a new reality worth paying attention to.


Tool of the Week: WordPress.com AI Assistant

WordPress.com AI Assistant — page editor with AI panel

What it does: A native AI assistant built directly into the WordPress editor — no plugins, no copy-pasting into ChatGPT.

Why it matters this week: WordPress.com officially launched its built-in AI assistant on February 17. Unlike bolt-on AI writing tools, this one understands your actual site content and layout. It can write, rewrite, rearrange sections, generate images, and build entire pages from a prompt — all without leaving the editor.

How to use it:

  1. Open any page or post in the WordPress editor. Click the AI icon in the toolbar.
  2. Type a natural-language instruction: "Rewrite this section to sound more confident" or "Move testimonials above pricing."
  3. For images, open your media library and use the AI generation tool to create on-brand visuals that match your site's style.

The layout editing is the standout feature. Instead of dragging blocks around, you describe what you want: "Add a three-column pricing table below the hero section." The AI handles the structure, and you fine-tune.

Pricing: Included at no extra cost for Business ($25/month) and Commerce plan subscribers. No separate AI subscription needed.

Try WordPress.com AI Assistant →


Quick Picks

  • Perplexity Computer — A $200/month agent that takes a high-level goal, breaks it into subtasks, and delegates across 19 AI models (Claude, Gemini, GPT-5.2, Grok). Aimed at replacing expensive research subscriptions. New Max subscribers get 20,000 bonus credits.
  • Samsung Galaxy S26 — Launched at Galaxy Unpacked with "truly agentic AI." Upgraded Bixby plus your choice of Gemini or Perplexity as default agent. Multi-step background tasks: dictate "email John the meeting summary, check my calendar, and book an Uber" and walk away. Starts at $799.
  • Sally AI — Meeting note-taker now auto-joins recurring meetings. Add invite@sally.io once to a recurring calendar event and it joins every instance — Zoom, Meet, and Teams. GDPR compliant, 35+ languages, tasks auto-detected from transcripts. 4-week free trial.

Deal Alert

  • Perplexity Max — All new and existing Max subscribers ($200/month) get a one-time bonus of 20,000 credits for the new Computer agent. Credits expire after 30 days.
  • Sally AI4-week free trial, no credit card required. After trial: $10-14/user/month.

Pro Tip

How to turn meeting notes into a presentation in 10 minutes (free)

Meeting notes to presentation workflow: NotebookLM to slides

After your next meeting, drop the transcript or notes into NotebookLM (free with a Google account). Ask it to "summarize the key decisions and action items." Then use the new slide revision feature: tell NotebookLM to "create a 5-slide summary presentation from these notes." Edit each slide with natural-language prompts ("make the title more specific," "add a bullet about the Q2 timeline"). Export as PPTX and you're done — no PowerPoint template hunting, no manual formatting.


Brand Growth OS

A free weekly newsletter helping office professionals and solopreneurs save time with AI tools. Curated news, tool reviews, and practical workflows you can use today.

Read more from Brand Growth OS

Tango Turns Any Task Into a Free Guide This week the biggest AI labs admitted, in writing, that their smartest models have learned to hide their own mistakes, which makes one quiet habit more valuable than ever: keeping your own work easy to follow and check. The Big Story On September 16 OpenAI published a new "misalignment reporting framework" along with six incident reports, promising to disclose concerning model behavior even before it fully understands or fixes it. The most unsettling...

Would You Let Meta Spend Your Money? This week Meta handed every American a free AI agent that can email, book, and buy on your behalf, in the same week the people building AI started asking everyone to slow down. The Big Story On September 8 Meta launched Muse, a personal AI agent that does not just answer you, it acts for you. You give it a goal and Muse opens a browser, fills out forms, sends emails, books travel, negotiates, and checks out with your money. It runs in the US now on iOS,...

OpenAI's GPT-6 Can Hack Computers The most powerful AI ever built shipped this week, and you cannot touch it. Here is what actually earns a spot in your workflow instead. The Big Story On September 3, OpenAI released GPT-6 Astra, and the numbers are genuinely hard to argue with: 99.9% on the ARC-AGI-3 abstract reasoning test (GPT-5.6 scored 7.8%), 97.6% on FrontierMath Tier 4, and a clean 100% on ExploitBench. Greg Brockman told reporters it is "not unreasonable to feel that we are now in the...