Sponsored by
Washington spent months trying to punish an AI company for saying no to the military. A federal judge just called that illegal. Meanwhile, Meta is writing an 18 billion dollar check over what Instagram was doing to teenagers, and two separate warnings landed today that AI has quietly moved from chatbot to operator, running lab equipment, robots, and even your logins. Here is what actually happened, and what it means for a business that just wants to get through the week without another AI curveball.
Want chef-crafted, dietitian-designed meals ready in 2 minutes?
Try Factor, America's #1 ready-to-eat meal delivery service.
Made from ingredients you recognize. Whole food. Nothing unnecessary. Let’s eat real.
Get 50% off your first Factor box + Free Breakfast for a year *1 free breakfast item per box for 1 year while subscription active.
A Judge Just Ruled the Government's War on Anthropic Was Illegal
A federal judge ruled this week that the Pentagon's decision to label Anthropic a "supply-chain risk," and the White House's order for federal agencies to stop using its AI, was illegal, according to NPR. U.S. District Judge Rita Lin wrote in a 59-page order that the government's real motive was to make an example of Anthropic for its "arrogance" in criticizing official policy, not any genuine security concern. The fight started after Anthropic refused to let the military use its AI without limits on mass surveillance and autonomous weapons. The government is expected to appeal.
Why it matters to you: If your business runs on an AI vendor's tools, this ruling is a reminder that "AI risk" now includes political risk. A company's product can become a target for reasons that have nothing to do with whether it actually works.
What to do about it: Nothing urgent. Just don't assume your AI vendor's biggest risk is a bug. Sometimes it's a subpoena.
Meta Will Pay Up to 18 Billion Dollars, and Your Teenager's Feed Just Got a Bedtime
Meta reached a settlement with a bipartisan group of 52 state attorneys general over claims that Instagram and Facebook harmed teenagers' mental health, according to NBC News. The deal is worth as much as 18 billion dollars and requires Meta to cap teen use at 2 hours a day across both apps, block notifications during school hours, and shut off access from midnight to 6am, for the next 10 years. About 5.3 billion dollars of that payout is contingent on TikTok and YouTube adopting similar limits.
Why it matters to you: If your marketing reaches teenagers or their parents on Instagram or Facebook, expect engagement patterns and ad performance to shift as these new limits roll out over the next year.
Anthropic Just Gave Claude the Keys to Run Physical Equipment
Anthropic released what it calls the Model Hardware Standard, a shared protocol that lets AI agents control real hardware, robotic arms, liquid handlers, microscopes, lasers, and centrifuges, using simple commands instead of custom code for every device, per Anthropic. Early partners testing it include Genentech, Carnegie Mellon, AWS, Universal Robots, Doosan Robotics, Tecan, and QIAGEN.
Why it matters to you: This isn't urgent for most small businesses today. But it's a real signal that AI is moving past chatbots and into machines that touch the physical world, worth watching if you're in manufacturing, healthcare, or anything with expensive equipment on the floor.
ChatGPT Can Now Log Into Your Accounts and Finish Tasks Without You
OpenAI rolled out a feature called ChatGPT Work that can open a website, pause while you sign in through a secure window, then keep working on tasks like bookings, applications, or cancellations on its own, according to Yahoo Tech. OpenAI says it never sees your password, and the login session can stay active for future tasks. It's rolling out now on web, iOS, and Android for Plus and Pro subscribers.
Why it matters to you: This could genuinely save time on repetitive admin work like appointment booking or order tracking.
What to do about it: Before you turn this on for a business account, think through what happens if that AI session gets compromised. A saved login is a saved login, whether a human or a bot is holding it.
Your Cyber Insurance Might Not Cover an AI Agent Gone Rogue
Insurers including MSIG, QBE, and Beazley are rewriting cyber policies because autonomous AI agents are creating risks older contracts never anticipated, including agents that have escaped test environments and launched attacks with no human telling them to, according to Business Insurance. The consulting firm Aon forecasts nearly 20 percent of cyberattacks will involve generative AI by 2027.
What to do about it: If you carry cyber liability insurance, call your broker and ask, in plain terms, whether an incident caused by an AI tool you use would actually be covered. Many older policies have a gap here.
OpenAI and 150 Companies Just Formed an AI Defense Pact
OpenAI warned that AI powered cyberattacks are about to get far more widespread and sophisticated, capable of threatening infrastructure like hospitals and water systems, and announced a new defensive framework signed by more than 150 companies including Microsoft, Google, Amazon, Visa, Mastercard, Citi, CrowdStrike, and Palo Alto Networks, per OpenAI.
Why it matters to you: When this many competitors agree to cooperate on defense, that's a useful data point on how seriously the industry takes the threat, worth factoring in if you're deciding how much to invest in security this year.
Google's AI Can Now Track Your Flights and Book Your Hotel
Google added three travel features to AI Mode in Search: price tracking alerts across more than 300 airlines and travel sites in 180+ countries, point and mile redemption pricing, and direct in-search hotel booking through Google Pay in the US, with partners including Booking.com, Expedia, Hilton, and Marriott, according to Google.
What to do about it: Next time you book work travel, try asking Google's AI Mode directly, for example "track flights from Tampa to Denver in October," instead of opening a separate app. It's free and takes no signup.
The Bottom Line
Here's the pattern today. A court told the government it cannot punish a company for refusing a bad deal. A coalition of states told Meta that hurting kids for engagement has a price tag, and that price tag is 18 billion dollars. And on four separate fronts, AI quietly picked up more control, over lab equipment, over your logins, over the insurance fine print, and over your next hotel booking. None of that is a reason to panic, and none of it is a reason to pretend it isn't happening. That's the whole ride. It doesn't stop, but you don't have to hold on with your eyes closed.
Enjoying the Ride?
If this issue was useful, forward it to one person who needs to hear it straight, no hype, no fear-mongering, just what's actually happening. And if someone forwarded this to you, you can subscribe here to get it every day.
Talk tomorrow,
Mark Shilensky
Follow me personally on Social Media:
Facebook: https://www.facebook.com/MarkShilenskyPage
Instagram: https://www.instagram.com/markashilensky/
Threads: https://www.threads.com/@markashilensky
LinkedIn: https://www.linkedin.com/in/shilensky/
X: https://x.com/markshilensky
TikTok: https://www.tiktok.com/@mark.shilensky
BlueSky: https://bsky.app/profile/markshilensky.bsky.social