AdvertiseFee Deal Desk

Alphabet (Google)

Alphabet (Google) (GOOGL) Extends Fairwind AI Access to Japan After September Launch

Google (GOOGL) will give Japanese agencies and critical infrastructure operators early access to advanced AI cyber defense models via Fairwind.

Be a creator
November 17, 2026, 04:18 PM UTC4 min read
AI SummaryAI
  • Google Cloud CEO Thomas Kurian confirmed advanced AI access for Japan's government on 2026-11-16
  • The Fairwind Program launched September 2 as a limited-access scheme for vetted government agencies
  • CodeMender AI agent automates vulnerability discovery, verification and remediation alongside Fairwind models
  • Jasjeet Sekhon left Google DeepMind after roughly eight months to join Anthropic

Fairwind AI Access Reaches Japan

Alphabet’s Google will give the Japanese government and domestic critical infrastructure operators early access to its most advanced AI models for cyber defense, widening a framework the company opened in September. Google Cloud chief executive Thomas Kurian confirmed the plan in an interview with the Japanese business daily Nikkei that was published on Monday, 2026-11-16. Kurian told the paper that Google is expanding the September framework so Japanese organizations can apply it against a growing volume of domestic cyberattacks. He positioned the early access as support for Japan’s cyber defenses at a moment when the country’s attack volume keeps rising.

The build-out runs through the Fairwind Program, a limited-access scheme announced on September 2 for government agencies and vetted partners. Its design principle is timing: defenders should reach advanced cyber-defense capabilities before attackers learn to abuse the same AI capabilities. Participants receive AI models built for cyber defense, usable together with CodeMender, an AI agent that helps patch software by automating the discovery, verification and remediation of vulnerabilities. Priority access lists government agencies, cyber-security authorities and critical-infrastructure operators in healthcare, telecoms, energy and finance. Access is conditional: participating organizations are screened and must confine model use to internal cyber-security, incident-response and penetration-testing teams. Google announced Fairwind as a proactive cyber-defense effort, giving defenders early entry to capabilities that would otherwise reach attackers first. The Japan work extends an earlier commitment. On April 13 Google said it would support the development of vulnerability-detection agents led by the AI Security Research Center (CREATE) of Japan’s National Institute of Information and Communications Technology, supplying technology that includes the Gemini model family to raise detection rates across automotive software, medical devices and internet-of-things equipment. Google has not named the Japanese agencies or operators taking part, and the terms of the expanded access were not disclosed. The equity market has treated the expansion as steady-state news rather than a catalyst. For Alphabet (GOOGL), a daily close above the nearest resistance level, not the Japan program, is what would change the technical picture.

Sekhon Leaves DeepMind for Anthropic

Jasjeet Sekhon, Google DeepMind’s chief strategy officer, has left the lab after roughly eight months in the role and joined Anthropic, the company behind the Claude model family. Anthropic confirmed the appointment through a spokesperson, and Sekhon announced the move in a post on X. His new mandate centres on what Anthropic calls “epistemic rigor”: research into how Claude distinguishes what it knows from what it does not, and how it quantifies uncertainty. Sekhon describes that work as an extension of his long-running research methodology. The appointment closes a short chapter at DeepMind. Demis Hassabis, Google DeepMind’s chief executive, personally announced Sekhon’s hire as chief strategy officer in March 2026, placing him across technical research, commercialization and policy in the lab’s strategy planning for artificial general intelligence. At the time, Sekhon said he had joined because he judged Google DeepMind the lab best positioned to advance AGI safely. That tenure lasted about eight months before the November departure. Sekhon’s background spans academia, quantitative finance and frontier AI research. He joined Bridgewater Associates in 2018, served as its chief scientist and head of AI, and co-founded AIA Labs in 2023 with Bridgewater co-chief investment officer Greg Jensen to apply AI and machine learning to financial market research and investing. In academia, he taught at Harvard, the University of California, Berkeley and Yale, where he held the Eugene Meyer chair. His research covers causal inference, statistics, machine learning and generative AI, with particular attention to building reliable causal judgments where signal-to-noise is low and usable data is scarce. The problem Anthropic has hired him for is concrete: large language models can produce confident, well-structured answers even without sufficient evidence, a failure mode known as hallucination. Epistemic rigor aims beyond cutting the share of wrong answers, toward models that weigh the strength of their evidence, recognize the limits of their own knowledge and state uncertainty when the basis for an answer is thin.

Primary sources

COINOTAG's editorial and research desk.

AI-Assisted

AI-generated, AI-reviewed, under COINOTAG editorial oversight.