:

GOOGLE QUIETLY SHUTS DOWN PROJECT MARINER AI AGENT

AI DESK2 MIN READ
WED, MAY 6, 2026

Google has shut down Project Mariner, its web-browsing AI agent designed to complete tasks on behalf of users, just months after showcasing the technology at Google I/O 2025.

Project Mariner, which Google demonstrated as a Chrome-based AI agent capable of autonomously handling web tasks, has been discontinued. The shutdown came quietly, with the company moving staff members off the project as it reassessed its AI agent strategy. The discontinuation marks a reversal from Google's public positioning of the technology. The company had highlighted Project Mariner at its flagship developer conference, signaling confidence in the browsing agent as a core part of its AI roadmap. According to reporting by WIRED, Google had already begun shifting resources away from the Project Mariner team nearly two months ago. The decision appeared to coincide with broader industry developments, including competitive pressure from similar web-browsing AI agents being developed by other companies. Project Mariner was designed to interpret user instructions and autonomously navigate websites to complete tasks—booking appointments, making purchases, or gathering information. The technology represented Google's effort to build practical applications for its AI capabilities beyond chat interfaces. The shutdown reflects the challenges in deploying autonomous web agents at scale. Issues around reliability, safety, and real-world viability may have contributed to Google's decision to halt the project rather than advance it further. Google has not made an official public statement about the discontinuation. The quiet shutdown contrasts with the company's high-profile unveiling of the technology, raising questions about how Google evaluates and prioritizes emerging AI projects going forward. The move suggests Google is recalibrating its AI strategy in response to competitive dynamics and practical constraints. While Project Mariner may be defunct, the underlying concept of autonomous web agents continues to attract attention across the AI industry, with other companies pursuing similar technologies.

■ MORE FROM THE AI DESK

Z.ai released GLM-5.3's weights on Hugging Face under a new license that requires large companies to undergo security review before hosting the model. The change marks a departure from the standard MIT license.

1H AGOAI Desk

Anthropic has introduced the Model Hardware Standard (MHS), a unified interface enabling AI agents to operate robotic arms, lab instruments, and other physical devices. Early testing shows integration time has dropped from weeks to hours.

1H AGOAI Desk

Open-weight AI companies—those releasing freely available models—are attracting major acquisition interest from tech giants. The trend reflects growing capital investment in the business model of distributing AI models at no cost.

3H AGOAI Desk

Google Deepmind has upgraded its Co-Scientist AI system to autonomously plan experiments, operate lab equipment, and publish scientific papers. The Gemini-based multi-agent platform demonstrated experimentally validated results across materials science, chemistry, and medical AI development.

3H AGOAI Desk

■ SUBSCRIBE TO THE DAILY BRIEF

ONE EMAIL, 5 STORIES, 06:00 UTC. UNSUBSCRIBE ANYTIME.