Codex Convergence: Workers Cross Job Boundaries as AI Adoption Surges

OpenAI's research reveals that workers are crossing job boundaries with AI, and this phenomenon has significant implications for the crypto labor market. The company tracked usage of its agentic AI tool Codex from August 2025 to June 2026 and found that non-developer token consumption surged 137-fold among individual users and 189-fold among groups.

By June 2026, Codex accounted for 99.8% of all weekly output tokens. A notable trend emerged: over 25% of Codex work performed by business-function employees involved engineering or coding tasks, with people in Legal, Finance, and Recruiting effectively crossing into technical territory that would have required hiring a developer just a year earlier.

OpenAI's April 2026 'AI Jobs Transition Framework' categorized occupations by vulnerability to automation and found that AI can theoretically handle around 90% of tasks in the highest-risk roles. However, actual usage was considerably lower, with workers engaging AI tools for less than a quarter of that theoretical potential.

The geographic divergence between Europe and the US is also noteworthy: Europe has lower shares of high-automation-risk employment compared to the US. This matters for crypto companies deciding where to locate talent and operations.