WordPress announces new AI building blocks for developers

WordPress.org has revealed its comprehensive approach to artificial intelligence integration through four interconnected building blocks designed to enhance the platform's capabilities. The announcement, made on July 17, 2025, establishes foundational infrastructure for AI features across the WordPress ecosystem.

According to James LePage, Engineering Director of AI at Automattic and Co-Lead of the WordPress Core AI Team, "We are taking a foundational approach to AI in WordPress, creating the necessary building blocks for plugin, developers, and core alike to create and explore impressive AI solutions."

The initiative encompasses four primary components. The PHP AI Client SDK provides unified interface capabilities for multiple AI providers while maintaining platform independence. WordPress sites can integrate with various AI services without requiring separate implementations for each provider. The system supports both WordPress-specific features and broader PHP community compatibility through dual Composer packages.

The Abilities API creates centralized registration for WordPress capabilities. Plugin developers can register functionality through standardized schemas and permission structures. LePage explained this component "creates a central registry where WordPress capabilities are discoverable and accessible, enabling AI agents and automation tools to understand and interact with everything WordPress can do."

Technical specifications include type-safe data handling and compositional UI materials. The API supports both human interfaces and machine interpretation through structured documentation. Registration examples demonstrate how plugins can expose functionality to AI systems while maintaining security boundaries.

The Model Context Protocol Adapter implements open standards for AI assistant communication. WordPress sites can operate as both MCP servers and clients, exposing capabilities to external AI systems while connecting to other MCP-compatible tools. The adapter leverages existing REST API infrastructure for HTTP transport and WP-CLI for stdio connections.

Security measures include application password authentication and capability-based authorization. The system maintains granular permission controls while enabling broader AI integration. Protocol compatibility ensures WordPress sites can connect with emerging AI standards without requiring core modifications.

Experiments plugin demonstrates integrated capabilities

The AI Experiments Plugin combines all building blocks into practical implementations. According to project documentation, the plugin serves as "WordPress's AI laboratory and a reference for developers building AI-powered features." Initial functionality includes demonstration features and development templates.

The plugin provides testing environments for different AI models and workflow experimentation. Users can evaluate AI capabilities within WordPress admin interfaces and editor experiences. Development teams can examine implementation patterns before building production features.

Future iterations may include core integration pathways. The project follows WordPress's canonical plugin methodology, where experimental features mature through community feedback before potential core inclusion. This approach allows rapid iteration while maintaining platform stability.

Industry context reveals growing AI integration

Recent developments show increased AI focus across WordPress hosting providers. WP Engine introduced Smart Search AI and AI-Powered Recommendations on July 15, 2025, targeting natural language search capabilities and personalized content suggestions. These implementations require separate infrastructure compared to WordPress.org's approach.

WordPress.com established partnerships with Perplexity AI in July 2024 for content discovery enhancement. Publishers can opt into revenue sharing when their content generates advertising revenue through AI-powered platforms. The partnership includes opt-out features for content creators who prefer maintaining stricter distribution control.

The WordPress.org approach differs by providing infrastructure rather than specific AI services. Developers choose their preferred AI providers while using standardized integration methods. This maintains platform independence and prevents vendor lock-in scenarios common with proprietary AI implementations.

Marketing implications extend beyond technical capabilities. PPC Land's reporting on Automattic's marketing challenges highlights how advanced technical features often receive limited promotion compared to their significance. Industry observers note patterns where WordPress releases sophisticated functionality without corresponding awareness campaigns.

Development timeline targets core integration

The building blocks follow WordPress's Phase 3 Collaboration roadmap. Real-time collaboration features could incorporate AI assistance through commenting systems and workflow management. Media library enhancements may include AI-powered organization and editing capabilities.

Site administration interfaces represent primary integration targets. AI agents could assist users through natural conversation for site management tasks. The MCP Adapter enables external AI assistants to discover and execute WordPress actions through standardized protocols.

LePage indicated community involvement remains essential for project direction. "If you're interested in helping build (some of) the future of the WordPress project, join in on the discussion in the comments," he stated on social media announcements.

The WordPress Core AI Team maintains active development across multiple repositories. The PHP AI Client project collaborates with broader PHP community initiatives. Abilities API development focuses on semantic extensibility and compositional UI frameworks.

Technical documentation emphasizes plugin developer adoption. Composer package distribution allows immediate implementation without waiting for core integration. Multiple plugins can utilize the same infrastructure without conflicts through dependency management.

According to project timelines, initial implementations target experimental features and community feedback. Production-ready features may emerge through iterative development and testing phases. Core integration decisions will depend on community adoption and technical maturity.

The announcement positions WordPress for emerging AI capabilities while maintaining platform independence. Technical infrastructure supports current AI technologies and anticipated future developments. Community-driven development ensures diverse use case coverage and sustainable implementation approaches.

Timeline

• July 17, 2025 - James LePage announces AI Building Blocks roadmap on social media and WordPress.org
• July 17, 2025 - Four technical articles published detailing PHP AI Client SDK, Abilities API, MCP Adapter, and AI Experiments Plugin
• July 15, 2025 - WP Engine introduces competing AI toolkit with Smart Search AI and AI-Powered Recommendations
• July 30, 2024 - WordPress.com partners with Perplexity AI for content discovery enhancement
• September 2024 - WordPress legal disputes impact community focus on development initiatives

Key Terms Explained

Vendor Lock-in Vendor lock-in occurs when a customer becomes dependent on a specific vendor's products or services, making it difficult or costly to switch to alternatives. In the context of WordPress AI integrations, this refers to situations where websites become tied to particular AI service providers through proprietary APIs or custom implementations. WordPress.org's building blocks approach specifically addresses this concern by providing standardized interfaces that work across multiple AI providers, allowing site owners to change AI services without rebuilding their entire infrastructure.

Canonical Plugin Methodology The canonical plugin methodology represents WordPress's strategic approach to feature development, where new functionality is first released as standalone plugins before potential integration into WordPress core. This methodology allows for extensive community testing, feedback collection, and iterative improvement without affecting the stability of the core platform. For AI features, this means experimental capabilities can be refined through real-world usage before becoming permanent platform components, reducing risk while maximizing innovation potential.

Model Context Protocol (MCP) Model Context Protocol is an open standard that defines how applications provide context and capabilities to large language models and AI assistants. In practical terms, MCP enables AI systems to understand what actions they can perform on behalf of users and how to execute those actions safely. For WordPress, MCP implementation means AI assistants like Claude or ChatGPT can discover and interact with website functionality through standardized communication methods, enabling natural language site management and content operations.

Semantic Extensibility Semantic extensibility refers to the ability of software systems to be extended with new functionality while maintaining meaningful, machine-readable descriptions of capabilities. This concept goes beyond simple plugin architecture by ensuring that new features can communicate their purpose, requirements, and interactions to both human users and automated systems. In WordPress AI context, semantic extensibility allows AI agents to understand not just what functions are available, but how they relate to user intentions and business objectives.

Compositional UI Frameworks Compositional UI frameworks enable developers to build user interfaces by combining smaller, reusable components rather than creating monolithic interface elements. This approach promotes consistency, reduces development time, and ensures maintainability across large applications. For WordPress AI features, compositional frameworks allow plugin developers to create interfaces that integrate seamlessly with core WordPress admin experiences while maintaining the flexibility to customize functionality for specific use cases.

Provider-Agnostic Architecture Provider-agnostic architecture describes system designs that function independently of specific service providers or vendors. In AI implementations, this means the same code and interfaces can work with different AI services like OpenAI, Google, Anthropic, or others without requiring separate development efforts for each provider. This architectural approach reduces development costs, increases flexibility, and prevents businesses from becoming dependent on single AI service providers for their website functionality.

REST API Infrastructure Representational State Transfer (REST) API infrastructure provides standardized methods for different software systems to communicate over the internet. REST APIs use standard HTTP methods and follow predictable patterns for data exchange, making them widely compatible and easy to implement. In WordPress AI context, REST infrastructure enables AI services to interact with WordPress sites through established protocols, ensuring reliable communication between AI systems and website functionality regardless of hosting environment or technical configuration.

Capability-Based Authorization Capability-based authorization is a security model where access to system functions is determined by specific permissions rather than broad user roles. Instead of granting general administrative access, this approach assigns granular capabilities like "edit posts," "manage users," or "modify themes" to individual users or systems. For AI integrations, capability-based authorization ensures that AI agents can only perform actions they're explicitly permitted to execute, maintaining security while enabling powerful automation features.

Multimodal AI Operations Multimodal AI operations involve artificial intelligence systems that can process and generate multiple types of content simultaneously, such as text, images, audio, and video. These capabilities enable more sophisticated interactions where AI can understand context from various input types and provide appropriate responses across different media formats. In WordPress applications, multimodal operations could enable AI assistants to help with content creation that involves both written copy and visual elements, or to analyze website performance across multiple content types.

Protocol Adapter Pattern The protocol adapter pattern is a software design approach that enables systems built for one communication standard to work with different protocols or interfaces. Adapters translate between different communication methods without requiring changes to the underlying system functionality. WordPress's MCP Adapter exemplifies this pattern by translating WordPress capabilities into MCP-compatible formats, allowing the same WordPress functions to work with any AI system that supports MCP standards while maintaining compatibility with future protocol developments.

Summary

Who: James LePage, Engineering Director of AI at Automattic and Co-Lead of the WordPress Core AI Team, announced the initiative alongside the WordPress Core AI development team.

What: WordPress.org revealed four AI building blocks: PHP AI Client SDK for unified AI provider integration, Abilities API for capability registration, MCP Adapter for AI assistant communication, and AI Experiments Plugin for practical implementations.

When: The announcement occurred on July 17, 2025, with technical documentation published simultaneously across WordPress.org AI team blog.

Where: The initiative spans the global WordPress ecosystem through WordPress.org infrastructure, targeting over 40% of websites worldwide that utilize WordPress.

Why: WordPress aims to democratize AI capabilities while maintaining platform independence, preventing vendor lock-in, and providing foundational infrastructure for diverse AI implementations across the ecosystem.