OpenAI NewsJul 18, 2024, 08:00 UTC+8

New compliance and administrative tools for ChatGPT Enterprise

OpenAI on July 18, 2024 launched an Enterprise Compliance API, eight third-party compliance integrations, and SCIM user management for ChatGPT Enterprise. The post confirms timestamped exports for conversations, files, GPT configs, memories, and users, plus support for Okta, Microsoft Entra ID, Google Workspace, and Ping; details on “Expanded GPT controls” are not disclosed in the provided body.

Read the original ↗Export Markdown

Take

OpenAI is selling less intelligence here and more permission to deploy. The Enterprise Compliance API exports timestamped conversations, uploaded files, GPT configs, memories, and user records, then plugs into eight compliance vendors including Forcepoint, Microsoft Purview, Netskope, and Palo Alto Networks. That is enterprise procurement language, not demo-day AI language.

This smells like OpenAI paying down the boring infrastructure debt that Microsoft and Salesforce customers already expect. Model quality gets a pilot started; SCIM plus Okta, Microsoft Entra ID, Google Workspace, and Ping decide whether a 10,000-seat rollout survives security review. The provided body does not disclose details on Expanded GPT controls, and that is the gap: if controls stay at coarse workspace toggles, internal GPT sprawl still gives CISOs an easy veto.

Source: OpenAI News · openai.com
Published: Jul 18, 2024, 08:00 UTC+8 · Jul 18, 2024