Breaking: Six AI Companies Release Near-Identical Agents for Knowledge Workers in Just Four Months
In a stunning convergence that has reshaped the enterprise software landscape, six leading AI labs—Anthropic, Perplexity, Microsoft, OpenAI, Google, and Amazon—shipped nearly identical AI agents for knowledge workers within a four-month window, signaling a new phase in workplace automation.
The Great Agent Rush
Anthropic kicked off the frenzy on January 12 with Claude Cowork, using Claude Code's proven agentic harness. Three weeks later, an open-source plugin pack erased $285 billion from the SaaS index. Investors panicked.
Perplexity followed on February 25 with Computer, an orchestrator routing tasks across nineteen models. Microsoft announced Copilot Cowork on March 9, built on Claude through its deepening partnership with Anthropic. OpenAI rebuilt the Codex desktop app on April 16, adding computer use, plugins, memory, and scheduled automations. Greg Brockman called it 'a general agent harness that happens to write software.'
Google launched Gemini Enterprise Agent Platform at Cloud Next on April 22, with long-running agents and daily briefings. Amazon closed the cycle on April 28 with its Quick desktop app, complete with personal knowledge graph and cross-platform connectors.
Background: Why They All Built the Same Thing
Every lab watched developers fall in love with Claude Code. Anthropic proved that agentic harnesses on frontier models could ship real work. The question became: why limit this to developers?
Anthropic's answer was Cowork. Kate Jensen told CNBC, 'We believe every knowledge worker will feel about Cowork the way engineers now feel about Claude Code.' Microsoft, OpenAI, Perplexity, Google, and Amazon saw the same opportunity and sprinted.
The core pitch is identical across all six: an agent that works alongside the knowledge worker, reads local files, drives the browser, retains context across days, and delivers finished outputs rather than suggestions.
What This Means
Knowledge workers are not developers. The Claude Code adoption curve assumed an audience that lived inside a terminal, understood file systems, and could read error messages. The Cowork pitch asks marketing managers, finance analysts, and HR leads to do something engineers took two years to learn: delegate multi-step tasks, supervise agents, catch off-rails moments, approve actions, and trust outputs not produced keystroke by keystroke.
The gap is behavioral, not technical.
Early evidence is mixed. Microsoft's April earnings call disclosed 20 million paid Copilot subscribers, up from 15 million in January—a 33% jump in one quarter. Real momentum, but still under 5% of the 450 million commercial Microsoft 365 base. PwC committed in April to rolling out Cowork and Claude Code to hundreds of thousands of professionals worldwide, the largest enterprise cowork deployment announced so far.
The supply side has buyers. Whether those buyers can overcome the behavioral barrier will determine if this convergence becomes a revolution or a short-lived hype cycle.
Related Articles
- 10 Ways GitHub Uses Continuous AI to Turn Accessibility Feedback into Inclusion
- 10 Key Facts About Python 3.13.10's Latest Maintenance Release
- How to Defend Against npm Supply Chain Attacks: Lessons from the AntV Incident
- How to Leverage Flutter 3.41 for Faster Development and Predictable Releases
- GitHub's Reliability Journey: Overcoming Rapid Growth Challenges
- New Surgeon General Nominee Nicole Saphier: Blending Medical Expertise with Media Presence
- Git Documentation Gets Major Overhaul: New 'Data Model' Document Clarifies Core Concepts
- Rust Expands Mentorship Horizons with Outreachy Participation in 2026