Google Forms AI ‘Strike Team’ to Close AI Coding Gap with Anthropic

Tech giant Alphabet (GOOGL) has put together a dedicated “strike team” of researchers and engineers to improve its AI coding models, as it looks to automate more of its own coding and, eventually, its AI research. According to The Information, the company is aiming for an “AI takeoff,” in which AI systems can improve themselves. This push is partly driven by advances from Anthropic, whose coding tools are viewed internally as stronger than Google’s Gemini models.

As a result, this project is being treated as a top priority within the company, with both co-founder Sergey Brin and DeepMind CTO Koray Kavukcuoglu directly involved. In a recent memo, Brin urged teams to move faster in building “agentic” AI, which are systems that can handle complex, multi-step tasks like writing code. The long-term goal is to create AI that can take over more of the work currently done by engineers and researchers. For comparison, OpenAI is already using internal tools to help its teams generate code for AI experiments, which adds to the competitive pressure.

Because of this, Google is also changing how it trains and uses its AI models. Instead of focusing mostly on tools for outside users, the company is now prioritizing models that can write code for its own internal systems. This includes training models on Google’s private codebase, which is more useful for performance but cannot be released publicly. In addition, Google is pushing employees to use these tools more often. Right now, about 50% of Google’s code is written with AI, compared to nearly 100% at Anthropic. As a result, improving coding AI has become a major focus as Google works to catch up.

Are GOOGL Shares a Good Buy?

Turning to Wall Street, analysts have a Strong Buy consensus rating on GOOGL stock based on 26 Buys and five Holds assigned in the past three months. Furthermore, the average GOOGL price target of $386.31 per share implies 13.7% upside potential.