A bailout in a blazer

Plus: The rivals who never agree just co-signed a warning to Congress. Pay attention to what scared them.

June 07, 2026

Sponsored by

My fellow AI explorers

There's a special kind of unease that hits when the people building the rocket start shouting about the brakes. This week, the three biggest names in AI, who normally won't agree on lunch, co-signed the same letter to Congress warning that their own tech is lowering the barrier to bioweapons. Then, hours later, Trump floated the idea of Uncle Sam buying a stake in those very same companies. And down in Nashville, a leap of leopards somehow became the face of the data center revolt.

In today’s edition:

Lindy remembers what you forgot to write down.

Lindy sorts your inbox, drafts replies in your voice, and preps you before every meeting over iMessage. 400,000+ professionals use it. Try it free.

THE BIG STORY

The Rivals Agreed on Something. That's the Scary Part.

The CEOs of Anthropic, OpenAI, and Microsoft AI just did something they almost never do: agree. Dario Amodei, Sam Altman, and Mustafa Suleyman co-signed an open letter to Congress urging the government to screen the buying and selling of synthetic materials that could be used to create bioweapons. When competitors this fierce link arms, it's worth asking what scared them into the same room.

Here's what you need to know:

  • The ask is specific. The letter wants Congress to mandate screening for companies selling synthetic DNA and RNA, materials the authors argue could be weaponized with AI's help.

  • The industry is volunteering itself for regulation. Some of the manufacturers of these materials, like Twist Bioscience and Ansa Biotechnologies, also signed, signaling part of the industry actually wants the rules.

  • The barrier is eroding fast. The letter warns that the knowledge barriers which historically kept bad actors away from biological weapons may meaningfully erode as AI improves.

The uncomfortable backdrop here is speed. A Stanford study this year found generative AI reached 53% of the world's population in just three years, faster than the PC or the internet. And it's not theoretical: experts have found that publicly available AI models can already provide information on how to create biological weapons and how to spread them.

Lawmakers aren't starting from zero. Senators Tom Cotton and Amy Klobuchar introduced the Biosecurity Modernization and Innovation Act of 2026 in February, which would force sellers to screen both orders and customers. The letter is essentially the industry telling Congress to hurry up and pass it.

🔮 Prediction: This is the canary moment. When the labs lobby for their own guardrails, they're not being noble; they're getting ahead of liability and shaping the rules before someone else does. Expect "safety theater" accusations from critics and a wave of compliance startups built specifically to screen synthetic biology orders. The companies that win the next phase won't just be the smartest; they'll be the ones who look the most responsible while staying the fastest.

AI Data Centers

Should You Own a Piece of OpenAI? Trump Thinks Maybe.

The weirdest political alignment of the year just got weirder. Trump said Friday that the U.S. government may take direct equity stakes in AI giants like OpenAI, Anthropic, and xAI, effectively endorsing the populist logic Bernie Sanders had laid out days earlier. "You make them a partnership in this revolution," Trump told reporters. "It would be a beautiful thing."

Here's the lay of the land:

  • The base is in revolt. Steve Bannon's War Room has been running episode after episode, attacking AI companies over copyright theft, job destruction, and power concentrated in a handful of unelected technologists.

  • Altman may have lit the fuse himself. OpenAI's CEO had been privately pitching a government ownership stake to administration officials before Sanders went public, and outlined a formal proposal in April for a U.S. public wealth fund.

  • This is already a pattern. The Cato Institute estimates the Trump administration already holds stakes in roughly 20 private companies, spanning minerals, semiconductors, and quantum computing.

Now for the part nobody on stage wanted to say out loud.

The numbers are brutal. OpenAI's own documents project $44 billion in losses between 2023 and 2029, with a $14 billion loss in 2026 alone. xAI lost $6.4 billion on just $3.2 billion in revenue in 2025, spending $2 for every $1 it brought in. Anthropic is the lone partial exception, on pace for its first profitable quarter with annualized revenue approaching $47 billion, a 47-fold jump from early 2025.

Which raises the question Washington dodged: a stake in companies burning tens of billions a year isn't a sovereign wealth fund, and calling it one doesn't change what it is. It's a bailout in a blazer. Dressed to look like a smart investment. Palantir's Alex Karp put it bluntly this week, warning peers that "the momentum is on the side of people who want to nationalize" these labs.

🔮 Prediction: Nationalization talk won't die down. It'll escalate into the 2026 midterms as a rare left-right rallying cry. The labs will pivot hard on messaging, leaning into "national champion" framing to make a government stake feel like patriotism instead of a rescue. Watch Anthropic try to stay out of the bailout bucket by waving its profitability around. Being the only one not bleeding cash just became its strongest political asset.

30-Second AI Play

Pressure-Test Any Big Decision With a Two-Sided AI Debate

When the smartest people in a room all agree, that's exactly when you should get nervous. Here's a prompt workflow to force your AI tool to argue against itself before you commit to anything, like a hiring call, a pricing change, or a big purchase.

  1. Open Claude, ChatGPT, or Gemini and describe your decision in one paragraph. Be specific about the goal and the constraints.

  2. Prompt: "Argue forcefully FOR this decision. Give me the three strongest reasons, with the risks each one ignores."

  3. New message: "Now argue forcefully AGAINST it. Same format, three strongest reasons, and what each one overlooks."

  4. Finally: "You've seen both cases. Name the single assumption that, if wrong, breaks my whole plan. Then tell me the cheapest way to test that assumption this week."

Why it works: most people use AI to confirm what they already believe. This flips it into an adversary, surfacing the blind spot you're paying it to find.

The magic is step 4, isolating the one load-bearing assumption so you're testing reality cheaply instead of betting big on a hunch.

💡 Pro tip: Paste the output into a fresh chat and ask it to "steelman the opposite of your conclusion." If it can't, your decision is probably solid. If it can, you just dodged a bullet.

Other Relevant AI News!

🍟 McDonald's is quietly testing an AI drive-thru order-taker called ArchIQ at five stores, and a franchisee claims it's already handled over a million transactions, with 90% needing no human help.

🇫🇷 France is rolling out its own battlefield AI command system named Arcadia as a sovereign alternative to Palantir's Maven, and it'll debut at a NATO exercise in Poland this month.

🐆 A leap of leopards just joined the data center backlash, with the Nashville Zoo gathering 180,000 signatures to block a facility going up 50 yards from its animals.

🥂 While the public sours on AI, Washington's elite threw a black-tie gala, the AI Honors Gala, complete with a dancing robot, even as new polling showed AI losing every trust matchup against other institutions.

⚔️ At that same gala, the Army Secretary said AI is helping the military "exponentially accelerate the kill chain." This line that lands very differently right after the bioweapon warning and other recent events like Pope Leo’s “Magnifica Humanitas” encyclical that warns us about the dangers of AI weaponry.

Golden Nuggets

  • ⚠️ When archrivals co-sign the same safety letter, the smart money watches what they're scared of, not what they're building.

  • 🏛️ Trump and Bernie agreeing on government AI stakes tells you the bailout-vs-ownership fight is the real 2026 story.

  • 🐆 The backlash is no longer abstract. It's leopards, lawsuits, and your neighbors, and it's coming for the buildout.

Would love to hear your thoughts! Send me your thoughts by replying to this email (yes, I read them all :)

Until our next AI rendezvous,

Anthony | Founder of Uncover AI