BotBeat
...
← Back

> ▌

Google / AlphabetGoogle / Alphabet
POLICY & REGULATIONGoogle / Alphabet2026-03-03

Google Violates Own 14-Day Deprecation Policy for Gemini 3 Pro Preview Model

Key Takeaways

  • ▸Google announced Gemini 3 Pro Preview deprecation on February 26 with a March 9 shutdown date, providing only 11 days of notice instead of the promised 14 days
  • ▸Google's official documentation explicitly states that preview models receive at least two weeks' notice between deprecation and shutdown
  • ▸The policy violation may undermine developer trust in Google's AI platform commitments and affect willingness to build on preview features
Source:
Hacker Newshttps://news.ycombinator.com/item?id=47235969↗

Summary

Google has failed to adhere to its own stated deprecation policy for preview models, giving developers less than the promised 14-day notice before shutting down Gemini 3 Pro Preview. According to the company's official documentation, preview models are guaranteed at least two weeks' notice between deprecation announcement and shutdown. However, Google announced the deprecation of Gemini 3 Pro Preview on February 26, 2025, with a planned shutdown date of March 9, 2025—a span of just 11 days, three days short of the promised two-week window.

The discrepancy was highlighted by a developer in the community who pointed to both Google's documented policy on preview models and the official changelog entry confirming the truncated timeline. The 14-day deprecation policy exists in Google's Gemini API documentation as a commitment to developers building applications on top of these experimental models, allowing them adequate time to migrate to alternative solutions or updated model versions.

This violation raises concerns about Google's reliability as a platform provider for AI developers, particularly those building production systems that depend on stable API commitments. While preview models are inherently experimental and subject to changes, the existence of an explicit deprecation policy suggests Google recognizes the need for predictable transitions. The incident may prompt developers to question whether they can trust Google's documented policies for future model releases and deprecations, potentially affecting adoption of preview features in the Gemini ecosystem.

  • This incident highlights broader concerns about platform reliability and policy adherence among major AI providers
Large Language Models (LLMs)MLOps & InfrastructureMarket TrendsRegulation & Policy

More from Google / Alphabet

Google / AlphabetGoogle / Alphabet
RESEARCH

Deep Dive: Optimizing Sharded Matrix Multiplication on TPU with Pallas

2026-04-05
Google / AlphabetGoogle / Alphabet
INDUSTRY REPORT

Kaggle Hosts 37,000 AI-Generated Podcasts, Raising Questions About Content Authenticity

2026-04-04
Google / AlphabetGoogle / Alphabet
PRODUCT LAUNCH

Google Releases Gemma 4 with Client-Side WebGPU Support for On-Device Inference

2026-04-04

Comments

Suggested

OracleOracle
POLICY & REGULATION

AI Agents Promise to 'Run the Business'—But Who's Liable When Things Go Wrong?

2026-04-05
Google / AlphabetGoogle / Alphabet
RESEARCH

Deep Dive: Optimizing Sharded Matrix Multiplication on TPU with Pallas

2026-04-05
AnthropicAnthropic
POLICY & REGULATION

Anthropic Explores AI's Role in Autonomous Weapons Policy with Pentagon Discussion

2026-04-05
← Back to news
© 2026 BotBeat
AboutPrivacy PolicyTerms of ServiceContact Us