OpenAI GPT-4o, o1/o3, and Operator
by Nick Clark | Published April 25, 2026
OpenAI operates GPT-4o, o1/o3 reasoning models, and Operator for agentic browser automation. Architectural element above OpenAI's platform — admissibility-as-skill-router with credentialed cross-jurisdiction operation — is what llm-skill-gating provides.
OpenAI Platform Reality
OpenAI operates GPT-4o conversational, o-series reasoning, Custom GPTs, Operator agentic browser, and emerging agent SDK products at substantial deployment scale. Platform-internal skill routing is operationally coherent.
Where Cross-Jurisdiction Friction Lives
Operator browser-automation and agentic operations face emerging EU AI Act, sector-specific compliance, and emerging adversarial-action concerns that platform-internal admissibility does not externalize structurally.
Architectural Substrate
Skill routing operates through credentialed admissibility; cross-jurisdiction operations admit through composite admissibility; lineage retention supports regulatory and incident audit.
OpenAI Trajectory
OpenAI gains regulatory-aligned architectural substrate ahead of pure-platform-replacement competitive pressure.
