Six major AI labs — Anthropic, Perplexity, Microsoft, OpenAI, Google, and Amazon — all launched AI agent products targeting knowledge workers within a four-month window. The convergence was triggered by Claude Code's success with developers, prompting every lab to ask why agentic harnesses should stay developer-only. The core challenge is behavioral, not technical: knowledge workers lack the mental models developers built over years of terminal use. Early adoption signals are mixed — Microsoft's Copilot hit 20M paid subscribers (a 33% quarterly jump) but still covers under 5% of its commercial base. PwC announced the largest enterprise deployment so far. The real question is whether seat licenses convert to daily habits, and whether adoption follows viral consumer curves or slow enterprise change management timelines.
Sort: