WordPress.com has introduced write capabilities for its Model Context Protocol (MCP), enabling AI agents like Claude, ChatGPT, and Cursor to create, edit, and manage content on WordPress sites through natural conversation.

All AI actions require explicit user confirmation, with new posts and pages starting as drafts and changes to published content flagged live; deletions move to a 30-day trash window.

Users retain full control, with every action logged in the Activity Log and options to disable specific capabilities or tailor permissions per site.

Trust and governance are central, emphasizing human oversight and granular per-operation permissions as the default configuration.

The feature is designed to reduce manual work for both business and personal sites while ensuring actions are reversible and auditable.

TechCrunch’s Sarah Perez frames the story, with attribution to Techmeme/TechCrunch in the article’s context.

Press materials include additional contextual links and media references, while the core message centers on new write capabilities and safety considerations.

The development is placed within a broader tech-news landscape, citing industry items like a White House AI policy framework and media-company moves.

The feature enables bots to perform a range of updates to reduce manual work, with all changes logged and reversible to preserve user control.

Explicit user confirmation applies to all AI-made changes, with the Activity Log providing transparency.

WordPress powers nearly half of all websites, underscoring the potential impact of AI-assisted publishing at scale.

Future expansions may include e-commerce management and broader site-operations automation, with competitive pressure likely to drive similar features across platforms.