Koray Kavukcuoglu is taking charge of Google DeepMind at a moment when the unit is under pressure to keep Gemini competitive with OpenAI and Anthropic, according to CNBC. Kavukcuoglu, previously DeepMind’s chief technology officer and Google’s chief AI architect, is stepping up to become senior vice president, replacing DeepMind cofounder and CEO Demis Hassabis, who becomes chair. The move puts a long-tenured DeepMind executive at the center of Google’s frontier-model effort. CNBC reports that Kavukcuoglu joined DeepMind in 2012, before Google acquired the lab in 2014. In the new structure, he will report directly to Google CEO Sundar Pichai and oversee Gemini model development, Frontier AI research, and the Gemini app and developer teams, according to a statement from Pichai cited by CNBC. The competitive backdrop is the main reason the change matters. CNBC frames the appointment as Kavukcuoglu inheriting a race to catch OpenAI and Anthropic at the frontier, where the most advanced models set expectations for developers, enterprise buyers and investors. The outlet reports that OpenAI and Anthropic have released new systems in recent months that drew praise and concern for their capabilities, while Google has not unveiled a frontier model since early 2026. Google’s position looked stronger after Gemini 3 in November 2025. CNBC says analysts described that release as moving the frontier forward and exceeding what they expected from that generation of models. But CNBC reports that Anthropic’s Mythos model and OpenAI’s GPT-5.6 were subsequently lauded, while Google’s releases after Gemini 3.1 Pro in February have not challenged at the frontier. Coding is one pressure point. CNBC reports that Google formed an internal team called Code Strike to improve coding capabilities, which the outlet describes as crucial to the broader push toward artificial general intelligence. Morningstar senior equity analyst Malik Ahmed Khan told CNBC that Google had focused more on monetizable areas such as Search and multimodal work, and had missed the first clear “killer use case,” coding, where he said OpenAI and Anthropic are well ahead. Google pushed back on the idea that it had been distracted from its central mission. A company spokesperson told CNBC it would be inaccurate to say Google was diverted from solving “intelligence” before “everything else,” and said the path to artificial general intelligence will require progress across agents, coding, robotics and world models. The spokesperson also pointed CNBC to DeepMind releases across robotics, video and computer use, along with policy and research work. Outside analysts see the leadership change as a shift in emphasis. Ben Wood, chief analyst at CCS Insight, told CNBC that Kavukcuoglu’s goal will likely include closing gaps with Anthropic and OpenAI in some areas. Wood also said the promotion appears designed to move Google’s AI effort away from academic projects and toward frontier performance and developer tooling. For now, the evidence supports a narrower conclusion: Google is reorganizing DeepMind leadership around Kavukcuoglu while external analysts press the view that Gemini needs stronger frontier and coding performance. Whether that becomes a measurable change in model capability will depend on Google’s next releases, not the org chart alone. Who benefits: Google benefits if Kavukcuoglu’s broader remit produces a more focused Gemini roadmap. Developers could benefit if the shift leads to stronger coding tools, the area CNBC’s analysts singled out as a competitive weakness. Who's exposed: Google DeepMind is exposed to execution risk if future Gemini releases do not challenge at the frontier. OpenAI and Anthropic remain the reference points in CNBC’s account, especially around recent model releases and coding capability.