Back to today’s briefing

Story perspectives

Google Cuts Meta's Gemini Access Amid $462B Cloud Backlog

6/28/2026

34 4 Full Breakdown

Summary

Google Limits Meta

Google limited Meta Platforms' access to the Gemini AI model after Meta requested more computing power.
The restriction disrupted Meta's AI work and affected other Google Cloud customers.
Alphabet CFO Anat Ashkenazi said Google Cloud's backlog was $462 billion at the end of Q1, with half expected as revenue within 24 months.
Q1 capital expenditures totaled $35.7 billion, about 60% of which funded servers.
Free cash flow was $10.1 billion.

Summary

Meta Pursues Gemini

In September 2025 Meta began talks with Google Cloud to adopt Gemini AI models for its advertising.
The negotiations also examined fine-tuning Google’s open-source Gemma models using Meta’s advertising data.
Meta planned to embed Gemini-driven insights into its ad-targeting and measurement platforms.
The proposed partnership signaled Meta’s intent to boost ad relevance and performance through advanced generative AI.

History

Historical Context

In July 2019 Microsoft poured $1 billion into OpenAI, securing exclusive Azure rights to its GPT-3 large-language model. That exclusivity set a precedent for cloud-AI deals, and in 2024 Google imposed similar restrictions on Meta’s access to the Gemini model, echoing the earlier Microsoft-OpenAI partnership. The pattern intensifies competition between Azure, Google Cloud and other providers.

1 of 3

Google Limits Meta

Google limited Meta Platforms' access to the Gemini AI model after Meta requested more computing power.
The restriction disrupted Meta's AI work and affected other Google Cloud customers.
Alphabet CFO Anat Ashkenazi said Google Cloud's backlog was $462 billion at the end of Q1, with half expected as revenue within 24 months.
Q1 capital expenditures totaled $35.7 billion, about 60% of which funded servers.
Free cash flow was $10.1 billion.

1 / 3