The EU Just Put WhatsApp In Timeout

Wednesday Briefing — June 10, 2026

🎉 Good morning. Brussels woke up, looked at WhatsApp, and decided the group chat had too much landlord energy.

Today's brief is about the new AI platform war: not who has the smartest model, but who controls the hallway every model has to walk through.

Let's ride. 🤠

EU Forces Meta to Open WhatsApp; Anthropic Files for IPO | Jun 10, 2026 (3 min)

Today's episode hits the two ends of the AI stack: the EU forcing Meta to reopen WhatsApp's business API to rival AI bots, and Anthropic reportedly marching toward the public markets with a $47 billion annualized revenue run rate. Translation: distribution and capital are becoming the real moats.

Listen here: EU Forces Meta to Open WhatsApp; Anthropic Files for IPO | Jun 10, 2026

The European Union just ordered Meta to restore WhatsApp access for rival AI chatbot companies while regulators investigate whether Meta is choking off competition inside messaging. The Associated Press report via ABC News says the order is an interim measure, which is regulator-speak for: "we are not waiting three years while the market hardens around you."

This is a big deal because WhatsApp is not some cute side app. It is the private messaging layer for a giant chunk of the planet. If AI assistants are moving from websites into messaging threads, customer support flows, and business inboxes, then the API becomes the new storefront. Whoever controls that API can decide who gets foot traffic and who gets the velvet rope.

Meta's argument is basically: the EU is making us give paid WhatsApp Business access to the largest AI companies for free. That is not insane. If OpenAI, Perplexity, Anthropic, or a vertical AI startup wants to sit inside WhatsApp, Meta would very much like a meter running. But the EU's argument is the classic platform argument: once you become essential infrastructure, your toll booth starts looking less like a business model and more like a moat with a fake mustache.

The Verge notes Meta has to comply by mid-June or risk fines that can reach into the tens of billions. That is the part founders should actually care about. The fight is not "Meta versus Brussels." The fight is "will AI companies be allowed to reach users through dominant apps, or will every major app become its own little kingdom?"

The last decade was about app stores. The next one may be about messaging rails. Your chatbot is not just competing on intelligence anymore. It is competing for where it is allowed to live.

The detailed ones. The ones with examples and edge cases. Wispr Flow lets you speak them instead — clean, structured, ready to paste into any AI tool. Free on Mac, Windows, and iPhone.

Try Wispr Flow free

Anthropic launched Claude Fable 5 and Mythos 5, with Mythos positioned as the company's highest-end model and Fable as the broader release. The interesting bit is not just the benchmark marketing. It is that Anthropic is openly talking about phased availability and safeguards like a company that knows the model is powerful enough to create new problems.

At the same time, Barron's reports Anthropic could be part of the next wave of blockbuster AI IPOs, with a $47 billion annualized revenue run rate. That number sounds fake until you remember how quickly AI spending turned into the new enterprise software tax.

Why it matters: Anthropic is trying to do two hard things at once. It wants Wall Street to believe it is a generational business, while convincing enterprises and regulators it is still the responsible lab in the room. That is a narrow bridge. The revenue is loud. The safety posture has to be louder.

China is reportedly preparing a roughly $295 billion AI infrastructure plan over five years, focused on data centers, domestic suppliers, and national compute capacity. Sherwood framed it cleanly: China is building the network that can power its AI buildout without waiting on the U.S. chip stack.

This is the global version of "own the picks and shovels." Models get the press release. Data centers get the budget. If you believe AI is a national productivity race, then power, land, cooling, chips, networking, and financing are not back-office details. They are the product.

Why it matters: The AI cold war is not only about who has the flashiest chatbot. It is about who can keep the GPUs fed when everyone else is stuck in a permitting meeting.

Standard Bots raised a $200 million Series C at a $1 billion valuation, led by RoboStrategy, with backing from names including General Catalyst. The company says it is expanding its New York manufacturing footprint and is on pace to deliver a meaningful share of new U.S. industrial robot deployments.

That sentence sounds like boring industrial news until you remember the labor market is allergic to reshoring math. Everybody wants domestic manufacturing. Nobody wants to pay 2026 wages for every repetitive motion inside the plant.

Why it matters: The sexiest AI startup may not be another chat window. It may be the machine that makes a factory manager say, "Fine, I can actually run this line in Ohio."

ICEYE announced a EUR450 million Series F, valuing the radar satellite company above EUR10 billion. The pitch is simple and very not-small: governments and enterprises want near-real-time intelligence from space, and synthetic aperture radar works even when clouds, darkness, or bad weather ruin traditional imaging.

That matters because defense tech has crossed from niche procurement corner to venture-scale category. You can debate whether that is good for civilization over a long dinner. You cannot debate that the money is moving.

Why it matters: Space is becoming a data business with defense margins. ICEYE is one of the companies proving that "satellite imagery" is no longer a pretty map. It is a real-time operating layer for governments.

Business Insider reported LinkedIn's leaked roadmap includes creator subscriptions, a brand marketplace, and paid experiences. The platform seems to be staring at Substack, YouTube, and the consultant economy and asking the most LinkedIn question possible: "What if thought leadership had a checkout button?"

This could be cringey. It could also print. LinkedIn has the weirdest advantage in the creator economy: people are already in buying mode there. Nobody opens TikTok to approve a $12,000 retainer. Plenty of people open LinkedIn while pretending to work and accidentally buy a course.

Why it matters: The creator economy keeps splitting into entertainment creators and trust-based business creators. LinkedIn wants the second group. That is the smaller party, but the drinks are much more expensive.

NASA unveiled the Artemis III crew for a 2027 low-Earth-orbit mission that will test docking and life-support systems for future lunar landers, according to The Guardian. It is not the Moon landing yet. It is the expensive rehearsal before the expensive thing.

This is the part of space coverage people skip because it is less cinematic than "humans return to the Moon." But rehearsal missions are where the actual risk gets burned down. SpaceX, Blue Origin, NASA, and international partners all need the boring choreography to work before anyone gets a flag moment.

Why it matters: The moonshot economy is being assembled like a supply chain. Less poetry, more docking adapters.

Google made Gemini 3.5 Flash generally available, claiming it is 4x faster and strong enough to power the Gemini app, AI Mode, and enterprise use cases. Cheap and fast may matter more than genius if the product is being used millions of times a day.

Apple's WWDC AI updates did not exactly make Wall Street throw a parade. The market reaction was a reminder that "we are working on it" is no longer a strategy when OpenAI, Google, Meta, and Anthropic are shipping like caffeinated contractors.

Meta reportedly cut around 8,000 roles, with Business Insider saying managers and software engineers were hit hard as AI spending keeps climbing. The new org chart is simple: fewer middle layers, more compute bills.

Seattle's City Council unanimously passed an emergency pause on large new data centers. It is one of the cleanest examples yet of AI infrastructure crashing into city politics, power planning, and public trust.

Nintendo's June Direct brought another round of Switch-era nostalgia and new releases, with Insider Gaming tracking announcements across Rhythm Heaven, Onimusha, Dragon's Dogma, Stellar Blade, and more. The takeaway: Nintendo remains undefeated at selling adults their childhood in a new box.

CISA added fresh vulnerabilities to its Known Exploited Vulnerabilities catalog, and breach trackers kept the drumbeat going. AI gets the headlines, but plain old patch management is still where companies go to learn humility.

Ars Technica covered backlash to a giant data center plan that was cut in half after protests. The phrase "AI factory" sounds exciting until the factory wants your water, power, and patience.

Seattle's data center moratorium is not just a local zoning story. It is the future of AI infrastructure politics. Every executive wants more compute. Every city wants jobs. Every resident wants the lights to stay on and the water bill to not become a science experiment.

The AI industry has been selling magic. Now it has to show up with environmental impact reports, utility plans, tax promises, and a face that can survive a public comment meeting. Good luck, friends. The people who can explain transformer inference are not always the people you want defending a substation.

Meta's reported cuts are easy to read as just another Big Tech efficiency round. I think they are more specific than that. AI is letting companies ask a colder question: if one senior person plus a stack of agents can do what three teams used to do, which managers still need to exist?

That does not mean software engineers are cooked. It means the bar is moving from "can you complete tickets" to "can you own outcomes with leverage." The people who can direct tools, make judgment calls, and ship whole workflows get more dangerous. The people whose job is forwarding context into meetings have a bad quarter coming.

The funniest thing about a big Nintendo week is how little the company needs to sound like everyone else. While every software company is trying to turn itself into an AI platform, Nintendo can show up with a rhythm game, a remake, a weird little surprise, and grown adults immediately become eight years old again.

There is a business lesson hiding in the joy. Differentiation does not always mean adopting the new platform. Sometimes it means being the one company that knows exactly what emotional job it does for customers and refuses to over-explain it in a keynote.

Michael's saved-link pile this week has a theme: tiny teams using AI tools like operating leverage, not inspiration wallpaper. The good stuff is not "here are 50 prompts." It is operators showing the actual machine: voice agents, Codex-built sites, Claude reading the internet into Obsidian, Higgsfield creative workflows, LinkedIn ghostwriting systems, and agency math that gets interesting with one good VA instead of a 14-person Slack swamp.

One saved thread was the "$40k MRR, no employees" solo AI agency playbook. Another mapped the 5-client, one-VA agency math. Add OpenAI's Codex app/skills push and Anthropic-style workflow thinking, and the pattern is pretty obvious: the winners are not asking AI to replace the business. They are using AI to make the business small enough to stay profitable.

Why it has my attention: Every boring service business just got a leverage test. If you can turn research, production, QA, and distribution into a repeatable system, you do not need a huge team to make real money. You need a clear niche, proof of outcome, and the discipline to keep the machine from becoming a toy chest.

Rabbit hole to watch: The first wave of AI agencies sold prompts. The next wave will sell completed workflows with receipts: "here is the pipeline, here is the output, here is the dashboard, here is the ROI."

That's the briefing. Now go build something.

— Michael

P.S. If your business depends on one platform letting you through the door, today's WhatsApp story should make you sweat a little. Distribution is still the whole damn game.Until next time,

Michael Benatar