According to a report by the British 'Financial Times' on June 23, intelligence agencies of the 'Five Eye Alliance' jointly warned that although the West currently leads in the field of artificial intelligence (AI), this advantage may be short-lived. They described how Western adversaries could use AI to develop cyber-attacks that could overwhelm defense systems. Therefore, Western governments and companies must integrate AI into their cybersecurity systems as soon as possible; otherwise, they will face increasingly higher and avoidable risks.
The “Five Eyes Alliance” is the world’s largest intelligence organization, consisting of the United States, the United Kingdom, Canada, Australia, and New Zealand. Reports indicate that this warning jointly issued by the network security officials of the members of the “Five Eyes Alliance” follows its usual practice: demanding that companies bear most of the costs and risks associated with defending against attacks from adversaries.
The “Five Eyes Alliance” often offers recommendations such as reducing attack exposure and preparing for security incidents in a standardized manner. Western companies have been receiving such recommendations for decades.
But this time, the ‘Five Eye Alliance’ specifically called on Western companies to deploy AI models to strengthen network defenses. This actually indicates that an AI arms race has already emerged between defenders and attackers. However, it does not clearly state what measures the governments of Western countries will take to ensure the security of their own networks.
The United States Cyber Command will establish a Comprehensive Network Center and Joint Operations Center (ICC/JOC) at Fort Meade.
They warned: “Organizations that integrate AI tools into their security operations systems will be more secure. Those who delay will face increasing and avoidable risks.”
The report indicates that Western countries still hold a technological advantage in the field of AI at this stage. This includes the development of commercial AI, as well as the application of related technologies in military and intelligence reconnaissance areas. Overall, Western countries are ahead of Russia, China, and other nations in terms of progress.
But the Five Eyes Alliance warned in this joint statement that this lead advantage will not last long. “The time we have left is not several years, but just a few months.”
The statement acknowledges that advanced large models continue to improve in performance. On one hand, this can be beneficial for protecting critical infrastructure. On the other hand, it can also help attackers develop the ability to carry out proactive cyberattacks.
"Advanced artificial intelligence models may surpass current industry expectations, fundamentally reshaping the combat capabilities of both network attackers and defenders."
The statement did not name China, Russia, or any other countries, but aimed to call on Western governments and companies to take immediate action and prepare to counter the wave of high-intensity composite cyberattacks, which are designed using current artificial intelligence models and empowered by technology.
The report also states that Western cybersecurity professionals have actually encountered new threats posed by AI. These attacks are much more complex and can target a large number of targets simultaneously.
In May this year, Google's Threat Intelligence team stated that they successfully intercepted a special attack: the unknown attacker used artificial intelligence to discover a system vulnerability that had never been identified before, and exploited this vulnerability to carry out the attack.
Just one week before the announcement, the United States mandated that artificial intelligence company Anthropic restrict access to its most powerful artificial intelligence models by foreign personnel, stating that the leakage of such models could pose security risks.
It is worth mentioning that the claims of the 'Five Eye Alliance' reveal their deceptive tactics. The Financial Times reported in February that the US Pentagon is planning to collaborate with leading AI companies to develop AI tools for detecting and locating vulnerabilities in critical facilities and networks such as China's power grid, in preparation for potential military conflicts in the future.
In August last year, the China Internet Security Association released two cases where the US government used countries like Germany, South Korea, Singapore, and the Netherlands as stepping stones to carry out cyberattacks against China. The Chinese side responded that the US carried out these malicious cyber activities using its allies in Europe and around China as stepping stones. China has always argued that network security is a common challenge faced by all countries, and that it should be addressed through dialogue and cooperation.
