🎉 We shipped 1.0 of RegWatch - our API-first, agent-ready regulatory risk intelligence service

I have written about companies becoming hedge funds as the product cycles shrink.

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…

#TGIF #Episode-2 Agent C is back to saving the world created by unreliable agents.

I dont think Finland leadership realizes that the world has changed. The flavor of conversation is still stuck in the 90s when there is generational shift in attitudes around the world including India.

A strong enterprise learning loop across tech, process, and people will the moat in the next phase.

Amazing. @tylercowen estimates AI contribution to GDP growth rate to be 0.5%. We have a long way to go to deliver 2.7x productivity.

One other implication is that every company becomes hedge fund - placing multiple bets, and learning about where the opportunity is. This is where Carver comes into picture. We will give you that alpha to hedge in the product space.

#Wikipedia has governance and transparency issues, and lost sight of its mission. One user mentioned updated other pages such as Sanjeev Sanyal's to obviously drive a certain narrative. @grok summarize the governance issues at wikipedia

Logical next step to the growth of environments. Standardization will lead to rapid development and deployment of environments, and implicitly powerful agents with realworld impact #agents #devtools

Now that @a16z has officially discovered the regulatory intelligence category. You know what comes next - 100 startups funded by Thursday, launch videos promising one-click compliance by Friday, and a new primitive called "RegAgents" without which your risk team is basically operating by candlelight. Meanwhile, @carveragents will keep doing its boring old thing: actually working. We fully expect to be superhot next week and briefly tempted to declare ourselves "the new benchmark for regulated industries on earth, in orbit, and beyond." Jokes aside - if you're navigating complex markets and want risk and regulatory intelligence that already exists (and ships), let's talk.

AI Agent To Track Mastercard Rule Changes & Auto-update Policies - A Carver RegWatch Demo What happens when a card scheme updates its rules — and your internal compliance policies don't keep up? This demo shows a Carver RegWatch AI agent that monitors how Mastercard's Security Rules and Procedures (SPME) evolved across five releases between 2022 and 2025, then proposes the corresponding updates to a payment processor's internal compliance policies. The agent works in three stages: - Detects every section that changed between consecutive SPME releases - Classifies each change by materiality: breaking, substantive, clarifying, or cosmetic - Proposes targeted edits to the affected internal policy files In this walkthrough you'll see: ✅ A timeline view of all five SPME release transitions with severity breakdowns ✅ Per-change cards grouped by materiality, searchable by section, policy, or severity ✅ Side-by-side redlines — original Mastercard text on the left, proposed policy edits on the right ✅ Source PDF links that open at the exact cited page — no hunting through 200-page documents ✅ Track-changes rendering for compliance review workflows ✅ Low-confidence flags where the agent surfaces its own uncertainty for human review Across five releases: 214 proposed revisions, 15 breaking changes, 8 policy areas affected. This is what a compliance AI agent looks like when it's transparent about what it knows, what it proposes, and what it might have gotten wrong. --- Learn more: carveragents.ai Explore RegWatch: lnkd.in/daedFHRR Video link in comments --- *Source documents are public Mastercard publications archived via the Internet Archive. Policy baselines shown are synthetic and do not reflect any real production text.*What happens when a card scheme updates its rules — and your internal compliance policies don't keep up?

This week, three regulators acted in the same direction simultaneously. -OFAC designated 12 individuals tied to IRGC oil operations. -FinCEN issued Alert FIN-2026-Alert002 — mandatory reference in any IRGC-related SAR. -The FCA, Bank of England, and Treasury issued a joint statement on frontier AI cyber risk. Each one is an independent action item. Together, they describe a week where sanctions, financial crime, and AI governance converged on the same senior leadership team. Three different functions. Three different deadlines. One week. >The OFAC designations require immediate screening, not at your next scheduled cycle. >The FinCEN alert requires SAR template updates and training, now, not at quarter-end. >The FCA-BOE-Treasury statement requires a board-level conversation, not a forwarded PDF. This is what regulatory convergence looks like in practice. And why function-by-function monitoring isn't enough anymore. Which of these three landed on your desk this week. And did it have a clear owner? #RegulatoryConvergence #Compliance #Sanctions #OFAC #AML #AIGovernance #FinancialServices #RegTech #RiskManagement #CarverRegWatch #CarverAgents

Not just slow. They are not delivering. Too much oxbridge, and too little relevance and impact.

Thought-provoking line of analysis framing AI value capture as determined by scarcity of frontier intelligence vs scarcity of context

Aftenposten has done a huge disservice to their country. I remember meeting Norwegians in my grad school. They were warm, brilliant, and furthest from the way aftenposten is portraying them to be - entitled, racist, and ignorant.

Great presentation on how to think about the AI market.

When defence spending, sanctions enforcement, and AI governance move in the same week, it's not a coincidence. It's a posture. Carver reads it. So you don't miss it - carveragents.ai → The EU Foreign Affairs Council greenlit €90 billion for Ukraine with drone procurement at the centre — and separately floated expanding naval operations in the Red Sea and Strait of Hormuz. European defence posture is no longer reactive. It's being institutionalised. → OFAC designated 12 individuals and entities tied to IRGC oil operations under Executive Order 13224. Assets blocked, US person transactions prohibited. Run these names now — not after your next scheduled screening cycle. → FinCEN issued Alert FIN-2026-Alert002 with explicit red flags for IRGC-linked illicit finance — and specifically mandated review of digital asset transactions and blockchain ledgers. This is the first time blockchain monitoring has been this explicitly required in an IRGC-related alert. → Additional IRGC names were added to the SDN List under Executive Order 13886 — carrying secondary sanctions risk. Non-US businesses in the Middle East, Europe, and Asia are in scope. The extraterritorial reach here is not ambiguous. → The FCA, Bank of England, and Treasury issued a rare joint statement: frontier AI has fundamentally changed the cyber threat landscape. Boards must understand it, resource against it, and govern it. Delegation is no longer a defence. #Sanctions #AIGovernance #RegTech #GeopoliticalRisk #CarverAgents

Jevons paradox in action

Interesting dynamic developing against the AI companies. Outcomes optimal for the AI companies are being seen as suboptimal for the market and society. That misalignment will result in adversarial actions from the masses. We are are seeing early versions of the same