Google Gives Japan Early Access to Its Cyber Defence AI as Attacks Hit Record High

Google Cloud will give Japan's government and critical infrastructure providers early access to Gemini 4 Carbon for Cyber, its defence-focused AI model, as the country faces a record wave of cyberattacks.

Nov 18, 2026 - 17:07
 0  6
Google Gives Japan Early Access to Its Cyber Defence AI as Attacks Hit Record High
A night aerial view of Google's office tower in Tokyo, with the illuminated Google sign on the glass facade above busy city streets.

Google Gives Japan Early Access to Its Cyber Defence AI as Attacks Hit Record High

Google will give Japan's government and critical infrastructure providers early access to its latest cybersecurity artificial intelligence model, widening a programme that had until now been offered mostly to American agencies and security firms.

Thomas Kurian, chief executive of Google Cloud, announced the expansion to Nikkei Asia on Monday, November 16, 2026, during a visit to Japan to discuss the initiative with senior government officials. The move brings one of Google's most capable AI models into the hands of Japanese defenders at a moment when attacks on the country are running at a record pace.

What Japan is getting

Through the expansion, Japanese partners will receive early access to Gemini 4 Carbon for Cyber, a version of Google's Gemini 4 Carbon model tuned for cyberdefence work. The model is designed to help security teams find software vulnerabilities, analyse threats and respond to incidents faster than they could on their own.

In Japan, the model will be available to the National Center of Incident Readiness and Strategy for Cybersecurity, the Digital Agency, and companies in the financial and telecommunications sectors, according to the report.

Google has been widening access to Gemini 4 Carbon in stages. The model was opened to paying developers and AI Ultra subscribers earlier in November, but the cyber-focused version remains gated, with vetted defenders getting it first.

The Fairwind programme reaches beyond the US

The Japan expansion builds on Google's Fairwind Program, launched in September 2026. Fairwind gives vetted governments, national cyber authorities, critical infrastructure operators and security companies early access to advanced AI models for defensive purposes, before they reach the wider public. Google says the programme had more than 650 partners at launch.

Until now, that access had been extended mainly to US government agencies and security companies, Kurian said. Japan is the first major expansion of the programme beyond the United States to be made public, and it puts Tokyo's defenders in the same early-access group as their American counterparts.

Google is not alone in this approach. Rival Anthropic introduced a similar programme for cyber defenders in April, reflecting a wider industry shift: the most powerful AI models are now being handed to defenders first, on the reasoning that giving security teams a head start matters more than a simultaneous public release.

Why Japan is moving now

The timing is driven by the threat picture. Japan recorded 233 cyberattacks last month, a record, according to Trend Micro, and the government has been moving quickly to strengthen its defences.

On November 6, the Japanese cabinet decided to allocate 10 billion yen, about 62.6 million dollars, from fiscal 2026 budget reserve funds to apply AI to vulnerability detection and defence upgrades. The Google partnership fits directly into that push, giving Japanese institutions a frontier model to work with as they build out those defences.

Alongside the model access, Google is also rolling out Scan for Japan, a new automated vulnerability detection service being offered to between 20 and 50 critical government institutions. The service is intended to find weaknesses in government systems before attackers can exploit them.

A bigger bet on security

The expansion also shows how central security has become to Google's cloud business. The company's security infrastructure draws on technology from Wiz, the cloud security firm Google acquired in March for 32 billion dollars, its largest acquisition to date.

Google held a 16 percent share of the global cloud infrastructure market last quarter, ranking third behind Amazon and Microsoft, according to Synergy Research Group. Winning the trust of governments on security is a central part of its effort to close that gap, and the Japan deal is the latest sign of that strategy in action.

For Japan, the immediate question is how quickly its agencies and infrastructure providers can put the model to work. The attacks are already at record levels, and the government has made clear it does not intend to wait for the public release to start defending itself.

Sources

  • Nikkei Asia — Google to give Japan early access to frontier AI for cybersecurity (November 17, 2026)
  • el-fondo — Google grants Japan early access to cyber AI model (November 17, 2026)

What's Your Reaction?

Like Like 0
Dislike Dislike 0
Love Love 0
Funny Funny 0
Angry Angry 0
Sad Sad 0
Wow Wow 0
Nahid Joy I’m Nahid Hasan Joy, a technology writer, web developer, and digital enthusiast with a strong interest in the ever-changing world of technology.