OpenAI Launches Presence Service for Enterprise Agents as Model Competition Squeezes Margins
Updated
Updated · The Register · Jul 22
OpenAI Launches Presence Service for Enterprise Agents as Model Competition Squeezes Margins
3 articles · Updated · The Register · Jul 22
Summary
OpenAI has rolled out Presence as a consulting-led service that charges enterprises for on-site help deploying AI agents, extending beyond selling models alone.
The move targets a market where AI models are increasingly commoditized, pushing OpenAI toward higher-margin work in the deployment plumbing around enterprise agents.
Presence is aimed at production use cases such as customer support, billing and IT requests, tying agents into company data, policies and workflows with guardrails and human review.
The launch deepens OpenAI's shift into stickier enterprise offerings as competition narrows model-performance gaps and vendors look for revenue beyond core model access.
Can OpenAI's engineer-led AI model scale without creating costly vendor lock-in for businesses?
If 80% of AI agents go rogue, can OpenAI's new guardrails truly guarantee enterprise safety?
With the EU AI Act's deadline next month, who is liable when a guarded AI agent makes a costly mistake?
OpenAI Presence Launch: How Token Efficiency and Deep Integration Are Shaping the Enterprise AI Race
Overview
OpenAI has launched Presence, a new enterprise platform that marks a major step forward in AI automation for businesses. Presence empowers companies by deploying advanced AI agents that can automate key processes like customer support and sales. These agents connect directly to a company’s internal data and workflows, allowing them to operate with strong contextual awareness and efficiency. OpenAI is already using Presence internally to manage its AI phone support for English-speaking customers, providing a real-world example of its capabilities. This internal use demonstrates how Presence can streamline operations and improve customer service through intelligent automation.