When a veteran law enforcement official with thirty-five years under their belt starts talking about artificial intelligence, you listen. Andrew Kan Kai-yan is stepping down as Hong Kong’s deputy police commissioner, and his final exit interviews are packed with sobering reality checks. The core takeaway isn't just about traditional street policing anymore. It's about algorithms, generative synthetic media, and the quiet underground currents of modern state security.
Artificial intelligence is changing the criminal playbook. Law enforcement agencies across major jurisdictions are scrambling to catch up with an era where malicious actors can spin up sophisticated misinformation campaigns in seconds. Hong Kong is no exception. As local officials peer into the next half-decade, the intersection of advanced technology and public safety has turned into a major flashpoint.
The Reality of AI-Driven Disinformation and Soft Resistance
You cannot understand Hong Kong's current security posture without looking closely at what officials call "soft resistance." Street-level unrest may have receded following the implementation of sweeping local security legislation, but the ideological battlefield has shifted online. Underground networks have adapted. They use arts, culture, and digital platforms to spread subversive narratives.
Now, generative technology pours fuel on that fire. Creating believable deepfakes, automated social media bots, and mass propaganda used to require a dedicated army of human operators. Today, a small group of bad actors can scale up text and video generation with minimal effort. Kan noted that the force anticipates malicious groups weaponizing AI to churn out anti-government material and target critical infrastructure.
If you look at overseas attacks on corporate computer systems and public utilities, the writing is on the wall. Critical infrastructure protection is no longer just about physical fences and armed guards. It is about firewall resilience against autonomous cyber threats.
Fighting Algorithms with Algorithms
You don't defeat automated threats with old-school patrol tactics. The Hong Kong Police Force is turning to the very technology it fears to defend the city. Fighting AI with AI has become an operational necessity.
Automated text-screening tools and natural language processing models now help authorities monitor public online chatter to flag potential threats early. The race is structural. Criminals use machine learning to hide their tracks, while investigators deploy neural networks to map out digital footprints.
According to retiring officials, current legal frameworks—including Section 50 of the Safeguarding National Security Ordinance and the Protection of Critical Infrastructures Computer Systems Ordinance—already factor in electronic and computer-related threats. Yet, nobody is pretending the rulebook is finished. Policymakers have already signaled that entirely new laws might become necessary as generative tools evolve faster than legislative bodies can draft amendments.
The Four Pillars of Ongoing Security Vigilance
Technology is just one piece of a broader puzzle. Looking ahead over the next five years, local security leadership has outlined four primary categories demanding constant vigilance:
- Soft Resistance: Covert ideological messaging spread through cultural channels and online media to foster public resentment.
- Homegrown Extremism: Isolated actors radicalized through encrypted digital channels and decentralized forums.
- Foreign Interference: External political pressure and geopolitical friction impacting local stability.
- Absconder Collusion: Fugitives operating abroad who maintain financial networks or attempt to direct underground activities locally.
While authorities have managed to crack down on individuals handling illicit funds or donations for overseas fugitives, the digital nature of modern finance makes tracking money a relentless game of cat and mouse. Cryptocurrencies and decentralized payment rails allow bad actors to bypass traditional banking controls, forcing financial intelligence units to upgrade their monitoring systems.
What Comes Next for Digital Governance
Hong Kong Chief Executive John Lee Ka-chiu recently outlined plans for a comprehensive AI risk-governance strategy. This framework targets the intersection of criminal intent and machine learning, setting safety guidelines for autonomous agents.
The strategy is pragmatic. You cannot ban breakthrough technology, nor can you ignore its potential for disruption. Governments worldwide are realizing that AI sovereignty is national sovereignty. If your infrastructure relies on foreign systems or lacks defensive capabilities, you are exposed.
The message from the top brass is clear. The era of low-tech subversion is over. The coming years will test whether legal statutes and automated defense systems can outpace the relentless evolution of synthetic threats. Stop underestimating what code can do. The battle for modern security is fought in the data centers, and the stakes couldn't be higher.
