OpenAI

Enterprise AI · AI Agents · Customer Support Software

Launch

OpenAI launches Presence to deploy governed AI agents for enterprises

The rollout relies on OpenAI's own engineers handling deployment, blurring the line between software vendor and consulting firm.

• OpenAI introduced Presence on July 22, 2026, a managed enterprise platform for building, deploying, and continuously improving AI agents across voice and chat channels.
• Presence pairs OpenAI's models with company-set policies, guardrails, and escalation rules so agents can resolve billing issues, insurance claims, or IT requests and hand off to a human when needed.
• The product is not self-serve: it rolls out through limited general availability, with deployments run by OpenAI's Forward Deployed Engineers or a short list of consulting partners.
• OpenAI says Presence already runs its own English-language phone support line, resolving about 75% of inbound calls without human intervention.
• Named early users include BBVA testing Mexican voice support, SoftBank trialing Japanese-language service, and Insurance Australia Group piloting disaster-response claims agents.
• Trade press framed the launch as OpenAI moving toward consulting-style, 'boots on the ground' enterprise pricing rather than pure software licensing, as raw model capability becomes commoditized.
• Presence shows OpenAI chasing revenue beyond model access by owning the deployment and governance layer, the part of enterprise AI adoption companies have struggled most to trust.