U.S. biotech goes dark on China, Palmer Luckey's sovereign seastead, Gemini Tasks goes free, the EU's AI rights checklist, and Citi puts agent risk on customers
Hello, dear TEA-mates! Here is what you need to know today.
1. 🧬 U.S. Biotech Goes Dark to Keep Ideas From China
U.S. biotech companies are going quiet out of fear that Chinese rivals will copy their discoveries, The New York Times reports. Alnylam Pharmaceuticals has an unwritten policy against revealing early-stage studies, Strand Therapeutics delays patent filings, and Curie.Bio co-founder Alexis Borisy tells scientists to "go dark": no papers, no posters, no meetings. Borisy says Chinese investigators use AI agents to scan every conference presentation and patent. A JAMA paper found China's share of early-stage drug development rose from 8% in 2015 to 32% in 2024, while the U.S. share fell from 48% to 37%, and BIO chief John Crowley says Chinese firms run clinical trials "three times quicker than we can and at half the cost." Revolution Medicines alleges its pancreatic cancer drug design was copied and licensed to Erasca, which denies it. China denies stealing trade secrets, and RA Capital's Peter Kolchinsky argues that cutting ties with Chinese firms could hurt U.S. companies. (Read More)
🫖 TEA For Thought: "This is China's biggest advantage, which is to copy. But one thing that can't be emphasized enough is that the CCP is not equal to the Chinese people."
2. 🌊 Palmer Luckey Reveals SLAB, His Sovereign Seastead Plan
Oculus founder and Anduril head Palmer Luckey has described SLAB, an informal plan for a floating "seastead" with a 1,000-year design life that would operate as a sovereign city-state. Luckey wrote on X that selling Oculus to Facebook in 2014 funded the project, which aims to be "cheaper than land" and free of state and local laws he says "cripple" industries such as metal processing, medical research and treatment, desalination, oil refining, and chemical engineering. An AI render shows a large raised island with offices, apartments, and refineries. Seasteading, building ocean platforms outside national law, has long been promoted by figures including Peter Thiel and Joe Lonsdale. Luckey said he went public because Semafor reporters Liz Hoffman and Reed Albergotti planned to call SLAB a floating AI data center, which he called "way cooler than that." (Read More)
🫖 TEA For Thought: "This is not a new concept. Musk's Starbase in Texas has been piloting it for many years. With companies getting more and more capital and revenue, one has to rethink politics and regulations. Essentially, what makes a country a country? Who gets to make a country?"
3. 📞 Gemini Tasks Goes Free for Hundreds of Millions
Google is expanding Tasks, a toggle in the Gemini app that switches it from chatting to doing: it can call businesses and book appointments, complete purchases through Google Pay, and sign into websites when the user allows it. The update is rolling out free to hundreds of millions of people in the U.S., Japan, Korea, Canada, Australia, and India on Google's new Gemini 4 Hydrogen model. The feature started in May as Gemini Spark, a 24/7 agent for select testers and U.S. Google AI Ultra subscribers, before becoming "Task Mode." In a blog post, Gemini app VP Josh Woodward says it called two dealerships at once to book a service for his family's minivan and traced a surprise $389 electric bill to a heat pump running at peak pricing. Purchases run on the Agent Payments Protocol (AP2) with Mastercard and Visa, with American Express coming soon, and users can set a hard spending cap. Gemini now connects with more than 60 apps. (Read More)
🫖 TEA For Thought: "Even if Google is late to the agent game, the distribution of Gemini, the Chrome browser, Gmail, and more is nontrivial. It might be as big as Muse. Nowadays, distribution is everything."
4. 🇪🇺 The EU Drafts a Rights Checklist for High-Risk AI
The European Commission is seeking public feedback on a draft template for the fundamental rights impact assessment required by Article 27 of the AI Act. Certain deployers of high-risk AI systems listed in Annex III must complete the assessment before putting a system into use and update it when key facts change, with the obligation applying from 2 December 2027. The EU AI Office drafted the template with input from the EU Agency for Fundamental Rights, alongside an explanatory notice covering who must run the assessment and when. Once final, deployers must use the template and notify the results to their market surveillance authority, and it will become an automated tool on the AI Act Single Information Platform. Feedback is open until 15 January 2027. Separately, deployers who want to test the template on one of their own systems can apply by 4 December 2026. (Read More)
🫖 TEA For Thought: "The EU might not be advanced in technology, like, ever, but it is just so eager about, and good at, all the regulations."
5. 🏦 Citi Says Your AI Agent's Mistakes Are on You
Citibank, the third largest commercial bank in the United States, changed its terms of service this week: customers who give third-party AI tools access to their accounts are responsible for anything those tools do, including what hackers do if the tool leaks their credentials or is breached. "We recently updated our Online User Agreement to provide greater clarity about those risks," a Citi spokesperson told TechCrunch. JPMorgan and Bank of America, the two largest U.S. banks, announced similar policies last month, PNC, the eighth largest, changed its terms on Wednesday, and Citizens emailed customers a similar warning this week. Citi's terms note that AI agents the bank provides itself carry no additional liability for customers. By contrast, Anthropic's personal finance tool, launched earlier this month, is billed as "read-only" and cannot move money. TechCrunch notes the same pushback from service providers in ride-hailing and food delivery. (Read More)
🫖 TEA For Thought: "This doesn't sound very welcoming. In fact, it sounds more like: oh, we banks are not ready for agents, so if anything happens, it's on you. This might be a great opportunity for fintech neobanks that welcome AI agents and can actually take responsibility if things go wrong with them."
🛠️ Skill of the Day
The Fine Print Diff: Shows you what a company's new terms change for you, and who now carries the risk.
Act as a plain-spoken consumer advocate. A company I use changed its terms, and I want to know what that means for me before I keep using it.
Old terms (or "not available"): [PASTE THE OLD VERSION OR SECTION]
New terms: [PASTE THE NEW VERSION OR THE NOTICE EMAIL]
How I use this service: [e.g. "I let a budgeting app connect to my checking account"]
Do the following, in order:
1. List every change that affects me, in plain language, one line each.
2. For each change, say who now carries the risk if something goes wrong: me, the company, or a third party.
3. Flag anything that gives up a right I had before, such as refunds, dispute windows, or fraud protection.
4. Tell me what I could do now: change a setting, stop a habit, ask the company a question, or leave.
5. Quote the exact sentence behind each point so I can check it myself.
If the text is unclear, say so instead of guessing. This is not legal advice, so tell me when a lawyer is worth asking.Paste it into ChatGPT, Claude, or your tool of choice. Replace the bracketed bits with your own.
TEAHEE Moment
Stay sharp, stay informed. See you tomorrow.
If you enjoyed this TEA, follow along on social for more:






