The Next Wave of Enterprise AI

What It Covers

Enterprise AI enters a new phase defined by two competing pressures: interface evolution and cost management. OpenAI's Codex updates target non-technical knowledge workers, Microsoft launches seven in-house models optimized for cost efficiency, and a Trump AI executive order formalizes voluntary model-sharing with reduced pre-release windows.

Key Questions Answered

• •Knowledge Worker Adoption: Non-technical workers are adopting OpenAI's Codex at three times the rate of developers, with 5 million weekly active users. 72% produce artifacts like PDFs or spreadsheets weekly, and 50% now run multiple simultaneous tasks — a behavioral shift that lets one worker operate at the scale of a small team.
• •Codex Sites as Knowledge Work Primitive: Codex's new Sites feature converts any artifact into a shareable web app or interactive dashboard without traditional software development. Treat disposable web apps — revenue planners, event dashboards, product launch hubs — as a core knowledge work output alongside slide decks and spreadsheets, not as a coding exercise.
• •Microsoft's Cost Optimization Play: Microsoft's MAI Thinking One, a 1-trillion-parameter mixture-of-experts model, targets enterprise cost reduction rather than raw benchmark dominance. When tuned for McKinsey workflows, it outperformed GPT-4.5 on quality while costing 10x less — positioning frontier model customization as the enterprise cost management strategy for late 2026.
• •Token Cost Pressure Reshaping Enterprise Strategy: Anthropic's Mythos model is burning through millions of dollars in tokens rapidly, with Anthropic currently subsidizing usage. Uber has already imposed a $1,500 monthly per-employee token cap. Enterprises should begin modeling token consumption now and build cost governance frameworks before agentic workloads scale further.
• •AI Executive Order: Voluntary but Formalized: The Trump administration's signed AI executive order reduces pre-release model sharing windows from 90 to 30 days, keeps testing voluntary, and explicitly prohibits mandatory licensing regimes. The NSA leads model assessment. Critics across the political spectrum — from Steve Bannon to Bernie Sanders — view this as infrastructure for future mandatory regulation.