Platforms Compete for Enterprise
- Google and other hyperscalers are shifting from chat demos to embedding AI across enterprise workflows and security. - Google recently added Gemini-powered Ads Advisor safety features and extended agent partnerships with Salesforce and SAP. - These moves pair product upgrades with large infrastructure deals, increasing operational lock-in for enterprise customers ( ).
Google is pushing its artificial intelligence pitch deeper into corporate plumbing, tying new workplace agents to ad tools, customer software and cloud contracts. (salesforce.com) (storyboard18.com) (techcrunch.com) On April 22, Google added three Gemini-powered safety features to Ads Advisor inside Google Ads: real-time policy troubleshooting, 24/7 security monitoring and instant certifications. Google said the tools are designed to catch policy problems, watch for account risks and speed approval work that marketers usually handle manually. (storyboard18.com) (searchengineland.com) The same day, Google Cloud and Salesforce said they were expanding their partnership so AI agents can take actions across both companies’ systems. The companies said Salesforce customers will be able to use Google’s Gemini models in Agentforce, connect Salesforce data to Google BigQuery, and use Google Workspace inside Salesforce workflows. (salesforce.com) SAP and Google Cloud announced a parallel deal on April 22 aimed at marketing teams. SAP said Joule agents in SAP Customer Experience products will connect with Gemini on Google Cloud, with Gemini Enterprise serving as a hub for agents working across SAP and Google systems. (news.sap.com) (destinationcrm.com) These announcements land as cloud vendors move past chatbot demos and toward software that can complete jobs inside the systems companies already use. In practice, that means the AI model, the business app, the data warehouse and the security controls are increasingly being sold as one stack. (salesforce.com) (news.sap.com) (ciodive.com) Google is pairing those software hooks with large infrastructure commitments. TechCrunch reported on April 22 that Mira Murati’s Thinking Machines Lab signed a new multi-billion-dollar agreement to expand its use of Google Cloud infrastructure, including systems built around Nvidia’s GB300 chips; Google separately announced expanded AI infrastructure capacity for the startup. (techcrunch.com) (googlecloudpresscorner.com) Google also used Cloud Next 2026 to roll out a broader enterprise package, including what CIO Dive described as an “Agentic Data Cloud” meant to connect data systems for AI agents. CRN reported that Google introduced a Gemini Enterprise Agent Platform and a $750 million partner fund at the event. (ciodive.com) (crn.com) Rivals are making the same argument from different starting points. Microsoft, OpenAI, Salesforce and Google all announced new workplace agent products this week, with each company promising that its assistants can move from answering questions to handling tasks across files, meetings, customer records and internal tools. (msn.com 1) (msn.com 2) (salesforce.com) For customers, the tradeoff is becoming clearer: fewer handoffs between tools, but more dependence on whichever platform connects the model, the data and the workflow. Google’s latest moves show that the competition is no longer just about the smartest chatbot, but about who owns the systems where work gets done. (salesforce.com) (news.sap.com) (techcrunch.com)