OpenAI launches Presence, a deployable enterprise AI agent platform

OpenAI says the enterprise challenge is no longer whether AI agents can work, but making them reliable enough for production as products, policies, and user behavior shift. Presence answers that with a structured deployment process: each agent gets only the knowledge and system access it needs, the company sets the rules, and after launch the system flags gaps that Codex suggests fixes for — but only after teams test them against the live version.

The product emerged from years of deploying agents at customer sites, and every deployment feeds generalized insights back into OpenAI's research and product work. That tight loop helped OpenAI reduce human handoffs in its own support line by 15 percentage points in 10 days. Early adopters span banking, telecom, and insurance, suggesting the platform is meant for high-stakes, regulated workflows rather than simple chatbots.

The real test for Presence is whether enterprise customers will accept an AI agent that learns on the job, even if humans stay in the loop. OpenAI frames the guardrails and approval steps as giving companies control, but skepticism may remain about deploying agents into mission-critical settings. How fast limited general availability expands, and whether customers confirm the quality and safety data OpenAI cites, will likely determine take-up.