AI startup Anthropic just triggered a two-day global stock market rout that wiped $285 billion from global software, legal, and data services companies. The culprit: plugins that automate the exact work these firms sell.
That shift, from infrastructure provider to application layer competitor, triggered the sharpest selloff in software stocks since America’s April’s tariff crisis. Anthropic released plugins on January 30 for its Claude Cowork agent on GitHub that automate legal work—contract reviews, NDA triage, compliance checks, legal briefings. By Tuesday, February 3, more than $285 billion had been wiped from software, legal services, and IT firms across three continents as investors raced to dump shares with any exposure to AI disruption.
Thomson Reuters, the legal research giant behind Westlaw, saw its stock dropped 18%, erasing $8.2 billion in a single session—its steepest decline on record. The selloff was fueled by fears of “seat compression”, as investors bet that Anthropic’s automation will lead law firms to slash the number of paid software licenses they buy for human staff. Britain’s RELX (which owns LexisNexis) fell 14%, losing $11 billion. The Netherlands’ Wolters Kluwer shed 13%, down $6 billion.
By Wednesday morning, all three slid another 3% as the selloff spread to asset managers and data providers. A Goldman Sachs basket of U.S. software stocks fell 6%, the largest one-day drop since April. The iShares Expanded Tech-Software Sector ETF has now fallen 29% since the start of January—the worst two-month stretch since 2008.
The market reaction reflects a fundamental question now confronting professional services: can AI agents handle the same work faster and cheaper than the platforms currently charging annual licensing fees?
Damilare Odedina
The competitive shift happening in real time
The plugins extend Claude Cowork, an agent-based desktop application Anthropic launched January 12 that operates directly on user computers and executes multi-step workflows without repeated prompting. The company open-sourced 11 plugins on GitHub covering legal work, sales operations, marketing automation, financial analysis, and data processing.
What distinguishes this from earlier AI legal tools is the source. Startups like Harvey AI and Legora have offered similar automation for over two years. Harvey AI was valued at $5 billion in June. Legora raised funds at an $1.8 billion valuation in October. Both companies build applications on top of foundational models from providers like Anthropic.
Now the model provider is packaging the workflow tools directly. That puts Anthropic in potential competition with its own customers while also threatening the legacy platforms those AI startups aimed to replace.
“Selling started in legal software/data-adjacent names, including Experian, the London Stock Exchange Group, Thomson Reuters and LegalZoom, then broadened across the sector,” according to Axios.
Software sentiment in these circles has become “radioactive,” Anurag Rana of Bloomberg Intelligence told Axios. A separate research note from Jefferies called it the “worst ever.”
Indian IT companies absorbed heavy losses. Infosys dropped 7.1%, Tata Consultancy Services fell 6%, Wipro declined 5%. Japan’s NEC, Nomura Research Institute, and Fujitsu lost between 7% and 11%. London Stock Exchange Group plunged 13% Tuesday, then fell another 6% Wednesday.
What analysts are saying
Morgan Stanley analysts noted the shift in a research note on Thomson Reuters: “Anthropic launched new capabilities for its Cowork to the legal space, heightening competition. We view this as a sign of intensifying competition, and thus a potential negative.”
“Our concern is that the seat-compression and vibe coding narratives could set a ceiling on multiples,” Billy Fitzsimmons, analyst at Piper Sandler, wrote in a separate note.
Scott White, Anthropic’s head of product for enterprise, told Axios that Claude can “render interfaces directly within it,” which could “drive even more engagement and interactivity with all these other business systems.”
The company positions itself as a complement to existing software rather than a replacement. Stock markets spent Tuesday and Wednesday pricing in a different scenario—one where AI providers don’t stop at selling infrastructure but move up the stack to capture application revenue directly.
David Adubiina
