Briefly
- Google plans to spend as much as $185 billion this 12 months on AI and cloud infrastructure to energy the “agentic period,” its CEO mentioned.
- Sundar Pichai mentioned almost 75% of latest code at Google is now AI-generated and authorized by engineers.
- New offers with Citi, Pondering Machines, and a $750 million accomplice fund present Google’s plan to monetize agentic AI.
Google is making one in every of its biggest-ever bets on synthetic intelligence.
Talking Wednesday on the Google Cloud Subsequent occasion in Las Vegas, CEO Sundar Pichai mentioned the corporate plans to speculate between $175 billion and $185 billion in capital expenditures this 12 months—up from $31 billion in 2022—to construct the infrastructure wanted for what he referred to as the “agentic period” of AI.
“As we transfer into the agentic period, we’re taking this to the subsequent stage,” Pichai mentioned. “We’re making huge investments now and for the longer term.”
The spending surge highlights Google’s effort to compete with rivals, together with Microsoft, Amazon, and OpenAI, because the business shifts from chatbots to autonomous AI brokers able to finishing duties with restricted human oversight. In line with Pichai, Google is already utilizing these techniques internally.
“At this time, almost 75% of all new code at Google is AI-generated and authorized by engineers, up from 50% final fall,” he mentioned. “We are actually shifting to actually agentic workflows.”
Nonetheless, regardless of the push into agentic AI, Pichai emphasised that human engineers overview the AI-generated code. He mentioned Google can be utilizing AI to automate elements of its cybersecurity operations, serving to groups course of giant volumes of menace intelligence quicker and reply to dangers extra shortly.
“Every month, our groups obtain unstructured menace reviews at a scale that may take hundreds of hours to overview—a virtually inconceivable process,” he mentioned. “At this time, our safety operation heart brokers mechanically triage tens of hundreds of unstructured menace reviews every month by accelerating the extraction of vital intelligence and filtering out the noise. It’s lowered menace mitigation time by over 90%; we’re extra on the entrance foot than ever earlier than.”
Google additionally used Cloud Subsequent to point out the way it plans to show that spending into income. The tech large introduced a $750 million fund to assist its 120,000-member Google Cloud accomplice ecosystem construct and deploy agentic AI merchandise. The initiative contains engineering help, early entry to Gemini fashions, and incentives for firms equivalent to Accenture, Deloitte, and McKinsey & Firm.
Whereas Google used Cloud Subsequent to point out the way it plans to show its AI spending into income, different firms, together with Citi and Pondering Machines Lab, revealed how they’re utilizing Google’s infrastructure and AI instruments to launch new merchandise and practice frontier fashions.
Citi unveiled “Citi Sky,” an AI-powered wealth administration assistant for U.S. shoppers. On the similar time, Pondering Machines Lab mentioned it expanded its use of Google Cloud’s AI Hypercomputer to speed up AI analysis and mannequin coaching.
“One factor that’s tremendous clear: We’re firmly within the agentic Gemini period,” Pichai mentioned. “The dialog has gone from ‘Can we construct an agent?’ to ‘How will we handle hundreds of them?’”
Editor’s word: This story was up to date after publication to right the day of the occasion.
Day by day Debrief E-newsletter
Begin each day with the highest information tales proper now, plus unique options, a podcast, movies and extra.

