OpenClaw 2026.2.23 enhances security and adds AI features

OpenClaw has released version 2026.2.23, which focuses on enhancing security and integrating advanced AI features. This update introduces support for Claude Opus 4.6 and implements several security measures, including optional HTTP security headers to protect against man-in-the-middle attacks. Developers will benefit from improved session maintenance and disk-budget controls, which help prevent data leaks and storage overflows, making it a significant upgrade for users concerned about privacy and security.

The release also includes important changes to the agent workflow, such as a new default setting for the browser SSRF policy, which now operates in "trusted-network" mode. This requires users to explicitly configure their settings, but legacy configurations can be migrated using the "openclaw doctor –fix" command. With these updates, OpenClaw continues to enhance its developer tooling and API capabilities, ensuring that users can effectively leverage AI automation while maintaining robust security standards.