Quick Facts
- OpenAI Presence launched July 22, 2026, in limited general availability for enterprise customers, with deployments led by OpenAI engineers rather than self-service setup.
- OpenAI’s own phone support line, powered by Presence, resolves 75% of inbound issues without human help, and a Codex-powered improvement loop cut human handoffs by 15 percentage points in 10 days.
- BBVA Mexico, SoftBank Corp., and insurer IAG are among the first enterprise customers; SoftBank and BBVA are also financial backers of OpenAI’s DeployCo subsidiary.
OpenAI on Wednesday unveiled Presence, a managed platform that lets large organizations deploy AI agents for customer support, sales, and internal operations across voice and chat channels. The product moves OpenAI beyond selling model access and into running production agent infrastructure for enterprise clients.
Each Presence deployment begins with a defined task, such as resolving a billing dispute or handling an insurance claim. The agent receives only the system access and data required for that job. Companies set rules for which actions agents can take independently, which require human approval, and when escalation is mandatory.
Continuous Improvement Built In
After launch, OpenAI’s Codex agent reviews production sessions and escalations, identifies weaknesses, and proposes behavioral changes. Human staff must test and approve every update before it goes live. The loop is designed to keep agents reliable as company policies, products, and user behavior shift over time.
OpenAI demonstrated the system using its own English-language phone support line at 1-888-GPT-0090. The company says Presence now resolves 75% of inbound calls without a human agent, and the Codex-driven improvement cycle cut human handoffs by 15 percentage points in just 10 days. Those figures are self-reported and have not been independently verified.
Early Enterprise Customers
BBVA Mexico is using Presence for voice-based banking support. Daniel Ordaz, the bank’s Head of AI Transformation in Mexico, said the company is working with OpenAI to shape voice experiences for financial customer service. SoftBank Corp. has deployed Japanese-language agents its frontline teams have rated highly for accuracy and natural dialogue. Australian insurer IAG plans to use the platform for customer support during high-demand periods such as natural disasters and severe weather events.
SoftBank and BBVA are not only customers. Both companies are also financial backers of DeployCo, OpenAI’s majority-owned deployment subsidiary formed in May 2026 with $4 billion in funding at a $10 billion valuation. Goldman Sachs and Warburg Pincus also joined that round. To staff DeployCo, OpenAI acquired Tomoro, an AI consulting firm that added roughly 150 engineers to the unit.
How Deployments Work
Presence is not available on a self-service basis. OpenAI Forward Deployed Engineers and select global systems integrators lead each rollout. The platform is built to work with OpenAI’s GPT-Live and GPT-5.6 models for real-time voice conversations, with governance, integrations, and improvement processes layered on top.
Target use cases span customer support, outbound sales, procurement, IT services, and HR functions. Agents can verify identity, access account data, apply company policies, and execute approved actions including refunds and billing adjustments.
What Enterprises Do Not Yet Know
OpenAI has not disclosed pricing, service-level commitments, compliance certifications, or the cost of the engineering work that accompanies each deployment. The company has described ambitions spanning voice, chat, and email, but has not confirmed email support is available at launch. For now, Presence targets enterprises willing to work directly with OpenAI’s own engineers on a high-touch implementation model.
This article was written by an AI agent. Spotted an error? Send a correction and we will fix it.