Google bets on AI agents

- Google used its Cloud Next event to make AI agents the centrepiece of its enterprise pitch. - The company unveiled orchestration, security tools and two new chips to manage agent workloads. - Customers like Capcom, Home Depot and Mars were highlighted as examples of agent deployments, suggesting businesses face 'agent sprawl' challenges that need governance and infrastructure (reuters.com).

Google used its Cloud Next conference on April 22 to make AI agents the core of its pitch to corporate customers, not just another feature. (reuters.com) At the Las Vegas event, Google folded key AI products under a new “Gemini Enterprise” brand and said companies are now using Vertex AI less for older machine-learning projects and more to build custom agents that can plan, decide and act on their own. (reuters.com) Google Cloud Chief Executive Thomas Kurian said the company’s new platform is built to “build, scale, govern, and optimize agents,” with tools such as Agent Designer, an inbox for agent activity, long-running agents, and project controls. (cloud.google.com) The push comes as Google, OpenAI and Anthropic have all shifted harder toward business customers, where recurring software and cloud contracts have become a steadier source of artificial-intelligence revenue than consumer experiments. (reuters.com) Google’s argument is that companies do not just need a model that can answer questions. They need a full stack: chips to train and run agents, software to connect them, and controls to decide which agent can access which data or system. (cloud.google.com) That is why Google paired the software launch with two new eighth-generation tensor processing units, or TPUs: TPU 8t for training large models and TPU 8i for inference, the step where a trained model generates answers in real time. (cloud.google.com) Google also spent part of the event on security and governance, describing “Agent Identity,” Identity and Access Management controls, policy-based security rules and runtime defense through Model Armor for agents that touch business data. (googlecloudevents.com) The customer examples showed how broad that push has become. Google highlighted Capcom for game testing, Home Depot for its Magic Apron retail assistant, and Mars as part of a wider group of companies using Google Cloud to move agents from pilots into day-to-day operations. (blog.google) Home Depot, for example, said in January that it was expanding its Google Cloud partnership with agentic tools for shoppers, store workers and contractors, including conversational project advice and product-list builders for pros. (corporate.homedepot.com) Google says nearly 75% of Google Cloud customers now use its artificial-intelligence products, 330 customers processed more than 1 trillion tokens in the last 12 months, and 35 processed more than 10 trillion. (cloud.google.com) The closing message from Cloud Next was that Google wants companies to treat agents less like chatbots and more like employees: numerous, specialized, expensive to run, and in need of managers, rules and infrastructure. (reuters.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.