Anthropic is discovering that owning a platform layer comes with real friction. The company introduced paid Claude Code access to subscribers, then announced those same users will pay extra to integrate third-party tools like OpenClaw, creating a tiered monetization structure that splits revenue opportunities but also splits user patience. The timing matters: Claude Code's value proposition depends partly on tool integration, so charging twice for the same workflow feels like taxing users for using the product as designed. Meanwhile, Claude Code's security posture is being tested in real time. Hackers are circulating leaked Claude Code materials bundled with malware, which suggests either the tool's internals are more exposed than Anthropic disclosed or the leak is comprehensive enough to be weaponized beyond intellectual property theft. The UK government's simultaneous courtship of Anthropic to expand in London reads as a hedge against US regulatory friction and a play to anchor AI infrastructure outside American oversight. Anthropic gets leverage in three markets at once: it can negotiate with US regulators by pointing to UK commitments, extract subsidies or tax breaks from London, and diversify its political risk. The company is learning to play capital and jurisdiction against each other while its products are still being reverse-engineered and its pricing model is still being stress-tested by its own users.
Sloane Duvall