OpenAI has launched Auto-review, a new mode within Codex that enhances its capabilities for longer and safer autonomous workflows by reducing the number of required approvals. This update is part of OpenAI’s broader strategy to advance enterprise AI with developer-centric tools, positioning itself amid growing competition from rivals like Anthropic. Additionally, the enhancements to Codex include support for in-app browser use, allowing it to autonomously test and debug front-end applications by simulating user interactions, while also maintaining thread context for continuity in long-term tasks.

OpenAI: OpenAI is an AI research and development company focused on building advanced models and agentic tools like ChatGPT and Codex for coding and enterprise workflows. In 2026, it has emphasized accelerating practical AI adoption across industries, particularly in developer tools and autonomous processes. OpenAI recently introduced Auto-review as a new Codex mode to enable safer, longer-running autonomous workflows with fewer user approvals.
Auto-review: Auto-review is a new execution mode within OpenAI’s Codex coding agent designed for extended autonomous operations. It minimizes manual interventions during tasks like testing, builds, and automations by using a separate guardian agent to pre-check high-risk steps in context. This update enhances developer productivity by balancing speed, safety, and uninterrupted workflows.

Enterprise Focus: OpenAI is advancing enterprise AI with developer-centric tools amid competition from rivals like Anthropic.
Codex Enhancements: Codex now supports in-app browser use to autonomously test and debug front-end applications by simulating user interactions.
Automation Persistence: Codex maintains thread context to resume and schedule long-term tasks without losing prior information.