Japan is set to establish a framework aimed at countering potential cyber threats posed by a new artificial intelligence model from the U.S., known as Claude Mythos, which its developer, Anthropic Public Benefit Corp, has refrained from releasing publicly due to cybersecurity concerns. This decision follows a public-private joint meeting held on April 24 at the Financial Services Agency, where attendees unanimously agreed to form a working group to address the risks associated with Mythos.
Among the prominent figures present at the meeting were Finance Minister Satsuki Katayama, Bank of Japan Governor Kazuo Ueda, and the presidents of three major banks, alongside a senior executive from the Japan Exchange Group. Katayama later emphasized the severity of potential cyberattacks on financial infrastructure, stating, “The extent to which such issues are widespread in the financial industry is incomparable to that in other industries. Such attacks could have an immediate market impact and may even trigger a broader loss of credibility.”
Katayama characterized the threat posed by AI, particularly Mythos, as “a crisis that is already upon us.” The AI model’s capabilities to identify software or system vulnerabilities have raised alarms globally, as it can easily uncover critical weaknesses that have long evaded human experts. This has led to fears that Mythos could be weaponized for large-scale cyberattacks targeting essential financial infrastructure.
Countries, including the United States, have convened emergency meetings with banking executives to discuss the escalating threat. Anthropic’s decision to withhold the release of Mythos underscores the seriousness of these concerns, with the company citing potential misuse if the technology were to fall into the wrong hands.
A senior executive at a major Japanese bank expressed anxiety regarding the ramifications of AI-driven attacks, stating, “If we were hit by an attack and customer information were leaked, we might have no choice but to shut down our systems and conduct all transactions in cash.” The banking sector, already grappling with rising costs associated with anti-money laundering measures, is now under pressure to enhance its cybersecurity defenses. However, regional financial institutions with constrained resources worry they may struggle to keep pace with essential system upgrades.
The threat of AI-enabled cyberattacks extends beyond the financial sector. Disruptions to critical infrastructure—such as power plants, telecommunication networks, railways, or airports—could have dire consequences for economic activity and daily life. In May of last year, Japan enacted legislation targeting cyberattacks and aimed at reinforcing defenses for vital infrastructure. However, this law does not adequately account for the sophisticated methods enabled by advanced AI like Mythos.
Yoshimasa Hayashi, the communications minister, underscored the urgency of addressing the challenges posed by advanced AI in cyber warfare, stating that it is “an urgent issue.” In a proactive move, the ruling Liberal Democratic Party held an emergency meeting with representatives from Anthropic on April 20, during which the company reportedly demonstrated a willingness to collaborate on Japan’s cybersecurity concerns.
In response to the growing threat, the LDP has called on government agencies to develop more robust defensive frameworks, and efforts to build these systems are currently proceeding at a rapid pace. The collaboration between public and private sectors is seen as a crucial step in fortifying Japan against the evolving landscape of cyber threats, particularly those leveraging advanced AI technologies.
As the international community grapples with the implications of AI in cybersecurity, Japan’s response may serve as a blueprint for other nations facing similar risks. The emphasis on cooperation and rapid development of defensive measures highlights the urgency of securing not just financial systems, but also critical infrastructures that underpin national and global stability.
See also
Anthropic’s Claims of AI-Driven Cyberattacks Raise Industry Skepticism
Anthropic Reports AI-Driven Cyberattack Linked to Chinese Espionage
Quantum Computing Threatens Current Cryptography, Experts Seek Solutions
Anthropic’s Claude AI exploited in significant cyber-espionage operation
AI Poisoning Attacks Surge 40%: Businesses Face Growing Cybersecurity Risks
