Google LLC is reportedly in discussions with an artificial intelligence coding startup called Mechanize Inc. over a deal that would see it license some of its technology and hire a number of its most talented employees. Business Insider was the first to report the discussions, citing information from four people familiar with the conversations.

Mechanize was founded last year, saying that its overriding goal is to eventually create AI systems that can automate the work done by every single knowledge worker in the world. It’s currently focused on developing simulated virtual environments, evaluation benchmarks and specialized grading systems that can be used to test and train AI agents on various kinds of business tasks.

Earlier this year, the startup announced it had raised $9.1 million in funding from investors including former GitHub Chief Executive Nat Friedman, Stripe CEO Patrick Collison and the podcaster Dwarkesh Patel. Its CEO Tame Bisoglu is notable for having previously co-founded the AI research outfit Epoch Artificial Intelligence Inc. Business Insider said Google is currently discussing a sum of around $1.5 billion to strike a non-exclusive licensing agreement for Mechanize’s technology. As part of that deal, it would also hire some of Mechanize’s experts in model evaluation and development, the sources said.

It’s believed that Google is interested in using Mechanize’s technology to enhance the performance of its AI models, particularly when it comes to coding. That’s because Google’s coding agents are generally considered to be inferior to those of Anthropic PBC and OpenAI Group PBC, which have seen strong enterprise adoption with Claude Code and Codex, respectively.

However, Mechanize’s ambitions extend beyond coding. “Our current focus is software engineering, but our long-term goal is the full automation of valuable work across the economy,” it says on its rather threadbare website.

The discussions between Google and Mechanize highlight two key trends in the AI industry today. The first is that coding has emerged as one of the most lucrative applications of AI models, while the second is that big technology firms are willing to pay serious sums of money to secure the most promising AI developer talent.

Google has done this kind of thing before. Rather than buying AI startups outright, it structures the deals as hybrid transactions that enable it to secure their technology through licenses and poach some of their top talents. It’s a useful strategy as it gets around the antitrust scrutiny that comes with full acquisitions.

One of the most notable examples was Google’s acquihire of Windsurf’s most experienced AI engineers, and it also licensed the company’s tech. Former Windsurf CEO Varun Mohan now leads the development of Google’s agentic coding platform Antigravity. Two years ago, Google struck a similar deal with Character Technologies Inc. to re-hire its co-founder Noam Shazeer and obtain non-exclusive rights to its technology. Shazeer recently left Google again, however, joining OpenAI.

Image: SiliconANGLE/Dreamina