Coding race and enterprise push
- Reports say Anthropic and OpenAI are overtaking Google in the AI coding tools market. - Google launched new multi-source Deep Research agents to challenge them in enterprise workflows. - The contest shows coding-centred products and enterprise integrations are primary battlegrounds for frontier teams and recruiters. ( )
Google is losing ground in AI coding tools as Anthropic and OpenAI pull ahead, while Google is redirecting the fight toward enterprise research agents. (latimes.com) (blog.google) The Los Angeles Times reported on April 22 that Google leaders are worried rivals now offer coding products that are more effective and more popular with business customers. The report said Anthropic’s Claude and OpenAI’s tools have gained momentum in a market Google once looked well positioned to lead. (latimes.com) On the same day, Google introduced Deep Research and Deep Research Max, two agents built on Gemini 3.1 Pro that can plan multi-step research, gather information, and produce cited reports. Google said the tools are aimed at finance, life sciences, market research, and other enterprise workflows. (blog.google) (ai.google.dev) A coding assistant writes and edits software; a research agent pulls together documents, databases, and web results into a report. Google’s new pitch is that companies want both, but the larger contract may go to the vendor that can connect models to internal systems employees already use. (latimes.com) (cloud.google.com) Google’s new agents are designed to pull from more than the public web. The company said support for the Model Context Protocol lets Deep Research query private databases and third-party sources, and Dataconomy reported Google is working with FactSet, S&P, and PitchBook on financial-data integrations. (blog.google) (dataconomy.com) That enterprise push was broader than one product launch. At Google Cloud Next 2026, Google framed the event around the “Agentic Enterprise” and rolled out a Gemini Enterprise Agent Platform inside Vertex AI, alongside new security and customer-service agents. (cloud.google.com) (crn.com) (reuters.com) The recruiting battle is following the product battle. The Times reported that Google has been trying to respond internally as Anthropic and OpenAI gain attention for coding performance, a shift that affects which labs attract engineers building the next generation of developer tools. (latimes.com) Google is not conceding coding work; it is trying to widen the field. Its latest message is that the winner will not just autocomplete code, but also search company data, cite sources, and fit into the software stack large customers already pay for. (blog.google) (cloud.google.com)
