SYDNEY, 25 February 2026 – Anthropic has significantly expanded its footprint in the enterprise sector this week, unveiling a suite of “Claude Cowork” updates designed to integrate its AI directly into the daily workflows of office professionals. The move, which includes new connectors for major software platforms, comes as the company’s valuation reportedly reaches $380 billion, cementing its position as a primary challenger to OpenAI and traditional software vendors.
Latest Updates: Claude Cowork and Enterprise Expansion
On 24 February 2026, Anthropic launched a series of connectors and plugins for its Claude Cowork platform. These tools allow the AI to interface directly with Google Drive and other popular enterprise software, enabling Claude to perform tasks across design, human resources, and wealth management roles.
The announcement has sent ripples through the tech industry, with market analysts noting both the disruptive potential for legacy software and a recent “olive branch” to vendors. While previous updates triggered sell-offs in software stocks due to fears of displacement, the latest roll-out includes cooperative features that saw some software shares climb on Wednesday.
Security and “Distillation” Controversies
Anthropic has also moved aggressively into the cybersecurity space with the limited research preview of Claude Code Security. This tool scans codebases for vulnerabilities, leveraging the model’s training on vast repositories to identify risks faster than human developers.
However, the company is currently embroiled in a geopolitical dispute. Anthropic recently alleged that three Chinese AI developers engaged in large-scale “distillation attacks.” These campaigns reportedly involved over 16 million exchanges across 24,000 fraudulent accounts intended to illicitly extract Claude’s capabilities to train rival models.
Financial Growth and Market Impact
Anthropic’s financial trajectory has seen unprecedented acceleration over the past 14 months. According to recent industry reports, the company has scaled its revenue at a “miracle” pace:
| Period | Annualised Revenue (Estimated) |
|---|---|
| December 2024 | A$1.5 Billion ($1B USD) |
| July 2025 | A$6.1 Billion ($4B USD) |
| December 2025 | A$13.7 Billion ($9B USD) |
| February 2026 | A$21.3 Billion ($14B USD) |
Public Sentiment and Safety Concerns
Despite its commercial success, Anthropic faces internal and external scrutiny regarding its core mission of AI safety. While CEO Dario Amodei has expressed discomfort with AI companies “regulating themselves,” critics have labelled the company’s strategy as “safety theatre.”
Public insights suggest a growing tension between the company’s research-heavy origins and the “speed to win” required by the current AI race. This was highlighted by the recent departure of a prominent safety researcher who warned that the “world is in peril” as commercial interests begin to outweigh safety protocols.
Frequently Asked Questions
What is Claude Cowork?
Claude Cowork is Anthropic’s productivity tool designed for knowledge workers. It uses AI agents to perform specific job functions and can now connect to external data sources like Google Drive to assist with professional tasks in real-time.
What is a “distillation attack”?
A distillation attack occurs when a competitor uses a frontier model (like Claude) to generate massive amounts of data, which is then used to train a smaller, cheaper model to mimic the original’s performance. Anthropic recently blocked thousands of accounts involved in such activity.
Is Anthropic a public company?
No, Anthropic remains a private company. However, it has received significant investment from major tech firms and was recently valued at approximately $380 billion (USD) following its rapid revenue growth through 2025 and early 2026.
What is the Responsible Scaling Policy (RSP)?
The RSP is Anthropic’s framework for managing the risks associated with increasingly powerful AI models. It dictates the safety protocols and “containment” measures that must be met before the company proceeds to the next level of model training or release.