# OpenAI rolls out Sign in with ChatGPT as an identity layer for apps > Source: > Published: 2026-07-31 05:54:54+00:00 OpenAI began rolling out Sign in with ChatGPT on Friday, a move [Greg Brockman (@gdb)](https://x.com/gdb), OpenAI's co-founder and president, [described on X](https://x.com/gdb/status/2083061510180688298?s=46) as support for a wider network of third-party applications. The identity-provider service lets ChatGPT users create, connect or access accounts on participating sites with their existing OpenAI credentials. OpenAI's [help documentation](https://help.openai.com/en/articles/20001410-sign-in-with-chatgpt), updated on July 31st, says the beta is available globally to authenticated users, including members of Enterprise organizations. Airtable, GitLab, HubSpot, Notion, Supabase and Vercel are the first participating partners. OpenAI Academy and Codex Sites also support the login system. The rollout turns the ChatGPT account into an identity that can follow users into software built outside OpenAI's own products. Brockman's backing carries specific weight. Before co-founding OpenAI, he was Stripe's [chief technology officer](/article/rob-witoff-joins-coinbase-as-cto) and helped scale the payments platform from four employees to 250, according to [his personal site](https://gregbrockman.com/). His work at OpenAI has included its API, Codex and large-scale model training. Sign in with ChatGPT applies the same platform logic to identity: give developers a common account layer, then use that layer to connect users with applications. ### What partner apps receive The initial login handoff is narrow. OpenAI says participating applications receive a user's name, email address and profile picture, when one exists. Signing in does not independently expose ChatGPT conversations, memory, files, tokens, billing information or other account data. Applications can ask for more access through a separate permissions flow. Users or their organization administrators must approve those permissions independently. That distinction matters as OpenAI tries to make a ChatGPT account useful across applications without giving every participating service access to the material stored inside ChatGPT. OpenAI also collects the information required to authenticate the user and secure the transaction. Its documentation lists account identifiers, the selected workspace or organization, application and authorization details, IP addresses, device identifiers and security signals among the data that may be processed during sign-in. Enterprise administrators receive an additional control layer. Sign in with ChatGPT is enabled by default when an organization has not established an explicit policy, but administrators can disable it or restrict employees to an approved list of applications through OpenAI's [Global Admin Console](https://help.openai.com/en/articles/12289294). Existing allow and deny policies remain in effect. [Vaibhav Srivastav (@reach_vb)](https://twitter.com/reach_vb/status/2082911102904435078), who works on OpenAI's developer experience and community efforts, introduced the feature in the thread amplified by Brockman. OpenAI has not published adoption figures or developer terms for the identity service. ### OpenAI's platform bet Sign in with ChatGPT gives OpenAI a place at the beginning of a user's relationship with another application. That position has long been controlled by identity systems from the largest consumer technology platforms. It can reduce registration friction for developers while giving OpenAI visibility into which applications its users connect and which permissions they approve. The six launch partners show where OpenAI is concentrating first: workplace software, developer infrastructure, databases and application hosting. Those categories also overlap with the tools that AI coding agents need to inspect, modify and deploy software. A shared identity can eventually make those handoffs easier, subject to separate permission checks. OpenAI laid out the larger strategy in its [March 31st funding announcement](https://openai.com/index/accelerating-the-next-phase-ai/), when it said it was building a unified product spanning ChatGPT, Codex, browsing and agent capabilities. OpenAI said the $122 billion round valued it at $852 billion after the investment. It also claimed more than 900 million weekly ChatGPT users and over 50 million subscribers at that point. That audience is the asset behind the new login button. OpenAI can offer developers access to users who already have an account and may already pay for ChatGPT or use it through an employer. In return, OpenAI gains another route for extending ChatGPT and Codex into outside products. The beta begins with ordinary identity information and a limited partner list. Its strategic value will depend on whether developers adopt it for deeper agent workflows and whether users accept OpenAI as the account connecting those services. OpenAI's permission model leaves room for that expansion without granting broader access at the moment of login.