Firefox maker torches Google for building Prompt API into browser

2026-04-30 · Source: The Register: Enterprise Technology News and Analysis · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Advanced, medium

Summary

Mozilla has reiterated its opposition to Google's Prompt API, a new technology being tested in Chrome and Microsoft Edge that allows web pages to directly prompt a browser-provided language model like Google's Gemini Nano. The API aims to facilitate local AI inference, offering benefits such as enhanced security, faster response times, offline usage, and cost-effective AI integration. However, Mozilla's Jake Archibald argues that the Prompt API poses severe negative consequences for web interoperability, updatability, and neutrality. Concerns include the potential for Google's Nano model to become a de facto standard, forcing other browser vendors to license it, and the requirement for developers to agree to Google's Generative AI Prohibited Uses Policy, which imposes restrictions beyond legal requirements. Furthermore, a performance report indicates that both Chrome's Gemini Nano and Edge's Phi-4 mini-instruct models using the Prompt API exhibit high failure rates in generative and classification tasks, with 15.17% and 23.93% task failures respectively for Chrome.

Key takeaway

For CTOs and VPs of Engineering evaluating browser-based AI integration, you should critically assess the long-term implications of vendor-specific APIs like Google's Prompt API. Prioritize solutions that maintain web interoperability and avoid tying your platform to a single vendor's model or content policies. Consider the reported performance limitations of current browser-embedded models and explore alternative, more open approaches for AI model access to mitigate future compatibility and policy risks.

Key insights

Integrating vendor-specific AI models directly into browsers risks web interoperability and platform neutrality.

Principles

• Prompts are tightly coupled to models.
• Model-specific code paths create browser compatibility issues.
• Web APIs should not enforce vendor-specific content policies.

In practice

• Use cloud service APIs for hosted AI models.
• Access local models via JavaScript runtime frameworks or WebGPU.
• Evaluate AI model performance before integration.

Topics

• Prompt API
• Mozilla Opposition
• Google Chrome
• Gemini Nano
• Web Interoperability

Code references

• mozilla/standards-positions

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Software Engineer, AI Engineer, AI Architect

Related on AIssential

• Google brings Auto Browse and Skills to Chrome Enterprise - and a new 'Gemini Summary' — Chrome Enterprise integrates agent-like AI and IT controls to automate tasks and manage AI security risks within the browser.
• Partnering with Mozilla to improve Firefox’s security — AI models like Claude Opus 4.6 can autonomously find and help fix high-severity software vulnerabilities at accelerated speeds.
• We’re expanding Gemini in Chrome to users in Asia Pacific. — Gemini in Chrome integrates AI for personalized browsing, content summarization, and cross-app functionality.
• Any Headphones = Live Translation — AI adoption faces regulatory and ethical considerations while simultaneously advancing in practical applications and autonomous agents.
• Firefox 151 adds full AI opt-out controls for mobile users — Firefox 151 offers unprecedented mobile AI opt-out controls, distinguishing Mozilla's user-centric approach to integrated AI features.
• AI Roundup 149: Flash Forward — Deployment economics and interoperable "skills" are shaping the competitive AI landscape amid infrastructure challenges and geopolitical chip races.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Register: Enterprise Technology News and Analysis.