:

GEMMA 4 SILENT UPDATE FIXES BUGS, BOOSTS PERFORMANCE

INDUSTRY DESK1 MIN READ
THU, JUL 16, 2026

■ AI-SUMMARIZED FROM 1 SOURCE ▸ TIMELINE

Google released a stealth update to its open-source Gemma 4 AI model, addressing tool calling bugs and truncated response issues while improving performance on Nvidia Hopper GPUs.

The update arrived under the same version name, meaning users may not immediately recognize the changes. The fixes target two persistent issues: problems with tool calling functionality and responses that were being cut off prematurely. The update also includes performance optimizations specifically for Nvidia's Hopper GPU architecture, potentially improving inference speeds for users running Gemma 4 on compatible hardware. Google has not made an official announcement about the update, which was identified through version tracking. The stealth deployment approach allows the company to push fixes without major fanfare, though it may leave some users unaware of improvements available to them. Gemma 4 remains part of Google's open-source AI initiative, competing with other accessible large language models in the market. The fixes enhance the model's practical utility for developers implementing tool use and automated workflows.

■ SOURCES

The Decoder

■ SUMMARY WRITTEN BY AI FROM THE LINKS ABOVE

■ MORE FROM THE AI DESK

Tech companies are defaulting users into generative AI features, forcing them to manually disable unwanted tools. Critics argue opt-in should be the standard for sensitive AI capabilities.

JUST NOWAI Desk

1Password has launched a browser integration enabling Claude to access stored login credentials and complete multi-step tasks on users' behalf without exposing passwords to Anthropic's servers.

JUST NOWAI Desk

Meta's Oversight Board warned that leading AI models may be overly restricting user expression. The independent group is expanding its focus beyond Meta's platforms to scrutinize AI systems across the industry.

2H AGOAI Desk

Meta and other major tech companies are deploying AI for content moderation, but the approach has a fundamental blind spot: it cannot protect users from non-consensual content. The backlash to Meta's Muse Image tool illustrates the limitation.

2H AGOAI Desk

■ SUBSCRIBE TO THE DAILY BRIEF

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