Anthropic showcases a new ace in AI: Cowork targets the 95% white-collar market, predicting demand will far exceed Claude Code.

Anthropic's Chief Commercial Officer Paul Smith made it clear in an interview on April 1st (Wednesday) local time that the company's latest launch of the general AI office assistant, Claude Cowork, is poised to have far greater business value and market size than the previously highly anticipated developer-specific tool, Claude Code. This statement not only reveals Anthropic's judgment on the future enterprise market but also marks a milestone in AI applications transitioning from the professional technical development field to a broader office white-collar market. Analyzing from the perspective of user base and market capacity, Paul Smith pointed out that in the vast majority of large enterprises, software engineers and other technical personnel comprise only 2% to 5% of the total workforce, which means that even if Claude Code achieves a monopoly position among programmers, its ceiling remains limited. In contrast, Claude Cowork, positioned as a "universal office assistant," aims to serve over 95% of non-technical employees, handling a vast array of daily tasks including complex document audits, cross-application data transfer, and automated process planning. Since its preview version was released in January 2026, Cowork's market adoption rate has surpassed the company's record in just a few weeks. Its growth trajectory is significantly stronger than that of Claude Code, released in the same period last year, validating the indispensable role of general intelligent agents in improving overall productivity. In terms of technology implementation and user experience, Claude Cowork represents the highest level of Anthropic's intelligent agent architecture, completely moving away from the dependency on command-line interfaces (CLI) seen in Claude Code to adopting an intuitive graphical user interface. A more significant breakthrough lies in its deep "Computer Use" capability, enabling the tool not only to interact through APIs but also to execute physical-level operations such as clicks, typing, and cross-software navigation directly on operating systems like macOS, just like human employees. By autonomously planning multi-step tasks, Cowork can independently perform tedious tasks like extracting key terms from hundreds of unstructured PDF contracts and generating analysis reports, making its end-to-end task closure capability a valuable tool for high-value white-collar industries like law, finance, and administration to reduce costs and increase efficiency. However, Anthropic's success in becoming a leading company in the field of artificial intelligence and a strong competitor to OpenAI is largely attributed to the Claude Code project the company recently disclosed that the product's annual revenue has exceeded $2 billion. Through products like Cowork, Anthropic is committed to helping users simplify a broader range of workflows and replicating its success in the software developer field. At the same time, Anthropic is competing with companies like OpenAI and Google to build more advanced artificial intelligence systems to drive intelligent agents and other software. In this widely covered interview, Smith stated that he expects the pace of product releases to accelerate further. "I think the pace of innovation will speed up," he said. "It can be quite certain that model releases will continue, and the frequency will increase." However, amidst its rapid business expansion, Anthropic is currently embroiled in a severe public relations and technical security crisis. Just a day before Paul Smith outlined the company's grand business vision on March 31, 2026, due to a packaging error by an employee while publishing a software package to the npm repository, the complete source code of Claude Code was accidentally leaked, involving up to 512,000 lines of code. Reports indicate that this leak not only exposed Anthropic's core system architecture and complex private prompts for training intelligent agents but also revealed a blueprint for an unreleased assistant codenamed "Tamagotchi." Although the company quickly retracted the related package files, the code has been widely mirrored, undoubtedly providing competitors like OpenAI and Google with an opportunity to glimpse into its technical foundation and casting a shadow over the widespread adoption of Cowork regarding AI security governance. Furthermore, this leak occurred at a sensitive time for the company. Anthropic is currently in legal proceedings with the U.S. government over the Department of Defense's decision to designate it as a supply chain risk company, with previous disagreements between the two on artificial intelligence security measures. The company warned that this label could result in the loss of billions of dollars in revenue. However, Smith noted that some customers acknowledge Anthropic's "principled stance" in negotiations with the U.S. government. Reports indicate that while battling with the U.S. Department of Defense, Anthropic is also advancing plans to go public as early as this year. Smith declined to comment on the initial public offering timeline, but stated that Anthropic is "very confident in steadily progressing towards its established gross margin and profit goals." Following a $30 billion financing in February, the company now has a substantial cash reserve that can be used to purchase expensive chips, data centers, and talent required for competing in the global artificial intelligence race. "We are very satisfied with our current financial situation after the recent fundraising round," he expressed. It's worth noting that the capital markets' reaction to Anthropic's shift towards general intelligent agents remains intense. Industry analyses indicate that Cowork's strong entry is directly challenging Microsoft Copilot's leading position in the enterprise AI market and even causing fluctuations in the stock prices of consulting and legal services companies that rely heavily on manpower. As Cowork gradually expands globally from a preview version, Anthropic is clearly no longer content being a mere technology provider but is attempting to build a new ecosystem through Cowork that could replace or enhance human office collaboration processes, thus establishing a decisive position in the increasingly fierce commercialization race of artificial intelligence.