TLDRs
- DeepMind reportedly uses rival AI coding tools internally despite Google restrictions
- Antigravity platform aims to unify Google’s fragmented AI coding ecosystem
- Competition in AI coding is shifting toward developer platforms, not just models
- Google faces pressure as internal teams prefer external AI tools for complex coding tasks
Google is facing renewed scrutiny over its AI development strategy after reports revealed that its DeepMind division is actively using rival AI coding tools.
Despite company-wide efforts to streamline its artificial intelligence stack, internal teams are reportedly relying on external systems, including tools from competitors, highlighting growing fragmentation within Google’s developer ecosystem.
According to industry reports, Google’s own Gemini-based capabilities are currently distributed across multiple products and workflows. This dispersion has created overlapping functionality and unclear strategic priorities, especially in AI-assisted software development tools. The lack of a unified approach is increasingly seen as a weakness as competition intensifies.
DeepMind Turns to Claude Code
A key concern for investors is that parts of DeepMind have been granted exceptions to use Anthropic’s Claude Code, despite internal policies generally discouraging the use of rival AI systems. This move underscores a practical challenge: some Google engineers reportedly find competitor tools more effective for complex tasks such as large-scale refactoring and multi-file reasoning.
https://x.com/WesRoth/status/2046499177950330952
Claude Code is particularly valued for its ability to handle structured reasoning across extensive codebases, a capability that makes it appealing for advanced engineering work. This internal reliance on external solutions raises questions about whether Google’s own tools fully meet the needs of its most advanced research teams.
Antigravity Platform Push Accelerates
In response, Google is consolidating its AI coding efforts under a unified platform known as Antigravity. The platform is designed to reduce fragmentation by bringing together coding assistance, debugging, and agent-based workflows into a single environment.
Antigravity is positioned as an “agent-first” system, where AI tools are not just assistants but active participants in software development. These agents can plan tasks, execute code changes, and verify results across multiple environments including editors, terminals, and even browsers. Google hopes this approach will simplify workflows and reduce reliance on external tools.
Despite internal challenges, Alphabet has reported that AI systems are already responsible for generating roughly 50% of its new code, signaling the growing importance of automation in its engineering pipeline.
AI Competition Shifts to Platforms
The broader industry trend suggests that competition in AI is no longer just about model performance but about developer ecosystems. With Antigravity, Google is shifting toward a platform-centric strategy that supports multiple underlying models, including those from external providers.
This means developers may eventually focus less on which model powers their tools and more on which platform best integrates their workflow. By supporting interoperability with models from both Anthropic and OpenAI, Google is attempting to position Antigravity as a central hub for AI-driven software engineering.
However, this strategy also introduces complexity. While it may improve flexibility, it risks reinforcing dependence on rival models if Google’s internal systems do not outperform them consistently in real-world tasks.
DeepMind Expands AI Coding Focus
Adding to the strategic shift, DeepMind has reportedly formed a dedicated AI coding team. This move reflects growing demand for tools that can assist in writing, debugging, and maintaining increasingly complex software systems.
As enterprise adoption of AI coding tools accelerates, Google is under pressure to ensure its internal platforms can compete not only in capability but also in usability. The success of Antigravity may ultimately determine whether Google can regain full control over its AI development ecosystem.


