Google pushes AI agents
- Google unveiled enterprise AI agents and new ad tools aimed at automation and monetisation. - Google Ads also added App Consent Insights to show advertisers how consent affects app-campaign performance. - Those moves push advertisers toward more automated, privacy-aware campaigns and require clearer asset metadata and consent handling ( ).
Google is tying two businesses together at once: enterprise artificial intelligence software and advertising tools that depend on user consent. At its Cloud Next conference on April 22, Google cast AI agents as a core product, while Google Ads rolled out new app-consent diagnostics for marketers. (reuters.com; searchengineland.com) At Cloud Next in Las Vegas, Google introduced a Gemini Enterprise Agent Platform that combines Vertex AI model tools with agent integration, security and DevOps features. Google said the platform is meant to help companies build and manage large numbers of AI agents, and Chief Executive Sundar Pichai said organizations want to “build, use and scale agents.” (blog.google; blog.google; blog.google) Reuters reported that Alphabet used the conference to show investors how it plans to make money from artificial intelligence beyond consumer chatbots. The company’s pitch was that AI agents can automate workplace tasks inside enterprise software, a market where Google is trying to gain ground on Microsoft, OpenAI and Anthropic. (reuters.com) In Google Ads, the new App Consent Insights feature shows advertisers how many app users have granted consent signals and how that affects campaign measurement and performance. Search Engine Land reported that the tool includes a rating, live counts and a split between European Economic Area users and users elsewhere. (searchengineland.com) Consent is the permission an app gets before collecting or using certain data, and Google now requires app advertisers to send those signals through consent mode for apps. Google’s help pages say advertisers need the latest consent mode application programming interface and updated Google Analytics for Firebase software development kit integrations to pass those signals correctly. (support.google.com; support.google.com; support.google.com) That puts more weight on setup work that used to sit in the background. Advertisers now need cleaner metadata for creative assets, better software development kit implementation and consent banners that meet Google’s policy and local privacy rules before automation can work as intended. (searchengineland.com; support.google.com; support.google.com) Google framed both moves as infrastructure, not just new features. On the cloud side, it is selling companies the tools to deploy digital workers; on the ads side, it is giving marketers a dashboard to see how privacy choices change the data that automated campaigns can use. (blog.google; reuters.com; searchengineland.com) The through line is that Google wants more of its customers’ work to happen inside Google systems, whether that work is handled by an AI agent or an ad campaign. The company opened the week by selling automation, and closed the loop by showing advertisers exactly how much consented data those automated systems can still rely on. (reuters.com; blog.google; searchengineland.com)