OpenAI's new Presence puts AI agents on the phone and in the chat window for companies
OpenAI has unveiled Presence, a platform for companies to build and manage AI voice and chat agents; it already runs OpenAI's own phone support and resolves 75% of cases with no human help.
After years of selling access to its models, OpenAI now wants to hand companies the worker ready-assembled. It’s called Presence, and it’s the latest move from the maker of ChatGPT: a platform for putting artificial-intelligence agents to work handling customers — on the phone and in the chat window.
What is OpenAI Presence?
It’s a platform for companies to build, govern and improve voice and text agents. Each agent is scoped to a specific task — customer support, sales, internal IT requests — and the company sets the rules: what the agent can do, when it needs approval, and when it should hand the call to a human. Under the hood, a tool called Codex proposes improvements based on what happens in real production.
What does Presence already do?
This isn’t just theory. Presence already runs OpenAI’s own English-language phone support — it verifies who’s on the other end, uses account context and takes approved actions. Within weeks, the company says, it met or beat the quality benchmarks of frontline human support, resolving 75% of contacts with no help from a person at all.
Which companies are trying it?
Big names. Spain’s BBVA, Japanese giant SoftBank and airline group IAG — owner of British Airways and Iberia — are already exploring the tool. It’s another step in the AI-agent race, in the same month that Microsoft moved closer to Mistral to bolster Europe. The official details are on the OpenAI site.
By Oliver Grant
Image: Jernej Furman from Slovenia / Wikimedia Commons (CC BY 2.0)