New strategy needed to fight crime in a globalised world

 As AI grows central to economic and national security, controlling its most valuable technologies will become a battle against new forms of transnational economic crime.

From P Sundramoorthy

A New York Times investigation into how a Chinese technology company placed on a US blacklist continued to access advanced American artificial intelligence (AI) chips raises questions that go well beyond the intensifying rivalry between Washington and Beijing.

It also offers an important lesson in contemporary criminology: regulation is increasingly struggling to keep pace with sophisticated corporate networks, rapidly evolving technology and transnational crime.

The investigation reportedly found that Inspur, a Chinese technology company placed on the US Entity List, continued to access advanced American technology through its US subsidiary and international networks.

Billions of dollars’ worth of advanced technology was reportedly shipped to Southeast Asia, including computers containing sophisticated Nvidia chips, with some of this computing infrastructure subsequently linked to Chinese tech firms.

Whether particular transactions constituted criminal or regulatory violations is ultimately a matter for the relevant authorities and requires evidence of knowledge, intent and unlawful conduct. Suspicious corporate relationships or transactions should not automatically be equated with criminality.

Nevertheless, the case illustrates how easily regulatory controls can be challenged by today’s globalised economy.

Traditional approaches to crime control tend to focus on identifiable offenders, physical goods and clearly defined jurisdictions. Sophisticated economic crime, however, increasingly operates through networks of companies, subsidiaries, intermediaries, logistics providers and technology platforms spread across several countries.

A company can be blacklisted in one jurisdiction while its subsidiary continues operating elsewhere. Goods can be shipped to a third country, incorporated into another system and subsequently made available remotely to users in yet another jurisdiction.

This creates what criminologists describe as regulatory arbitrage, the exploitation of differences between laws, regulations and enforcement capabilities across jurisdictions.

This is particularly relevant to Southeast Asia.

The region is rapidly developing as a centre for semiconductor manufacturing, cloud computing and data centres. These developments bring many economic opportunities, but they also create new vulnerabilities. Infrastructure established for legitimate commercial purposes can potentially be exploited to circumvent restrictions imposed elsewhere.

This does not mean Southeast Asian countries should become suspicious of every foreign investment or technology transaction. Nor should legitimate businesses be subjected to excessive regulation simply because they operate in strategically sensitive industries.

The challenge is finding the right balance between economic openness, national security and effective crime prevention.

From a criminological perspective, the concept of capable guardianship is particularly useful. Crime becomes more likely when valuable targets are available but effective monitoring and controls are weak.

Advanced AI chips and computing capacity are now extraordinarily valuable strategic commodities. Yet the systems intended to control their movement remain heavily dependent on national borders and conventional trade mechanisms. This creates a significant enforcement gap.

The problem is even more complicated because AI technology does not necessarily have to cross a border physically. A chip may be manufactured in one country, installed in a computer in another, operated in a third and accessed remotely from somewhere else.

What is effectively being transferred may therefore no longer be the chip itself, but computational power.

This presents a major challenge for law enforcement and regulators.

Monitoring containers at ports and checking export documentation remain important but they are no longer sufficient. Authorities increasingly need to understand corporate ownership, beneficial ownership, supply chains, end-users, data-centre operations and patterns of remote access.

The concept of corporate crime is equally relevant. Modern corporate wrongdoing does not necessarily involve a traditional criminal organisation. It may involve otherwise legitimate companies operating through complex commercial structures that deliberately exploit weaknesses in regulatory systems.

The challenge for enforcement agencies is determining where legitimate commercial activity ends and deliberate regulatory circumvention begins.

For Malaysia, the implications should not be underestimated. Malaysia is positioning itself as an important regional hub for semiconductors, digital infrastructure and data centres. This presents substantial economic opportunities.

However, the country must also ensure that its technological infrastructure is not inadvertently used as a channel for circumventing international restrictions.

Malaysia therefore needs stronger corporate due diligence, beneficial-ownership transparency, end-user verification, supply-chain monitoring and international intelligence sharing. Enforcement agencies should also develop specialised expertise in tracing sophisticated technology transactions and identifying unusual corporate structures.

At the same time, regulation must remain proportionate. Excessive restrictions could discourage legitimate investment and undermine Malaysia’s competitiveness.

The objective should not be to criminalise international business but to ensure that companies operating in strategically sensitive sectors meet appropriate standards of transparency and accountability.

The larger lesson from the New York Times investigation is that the geography of crime is changing.

The criminals of tomorrow may not resemble the traditional image of offenders operating from warehouses, ports or clandestine locations. Some may operate through sophisticated corporate structures, digital platforms and international supply chains.

The commodities may not be drugs, weapons or stolen goods. They may be computer chips, computing capacity, data and artificial intelligence.

Consequently, crime control must evolve.

Blacklists and export restrictions remain important tools, but they cannot be effective if enforcement stops at the border or at the name of a particular company. Regulators must increasingly follow the ownership, money, technology, data and the ultimate user across jurisdictions.

The real challenge is therefore not simply to impose more restrictions but to develop smarter, faster and internationally coordinated guardianship.

As AI becomes increasingly central to economic and national security, controlling its most valuable technologies will also become a battle against new forms of transnational economic crime.

For Malaysia and the wider region, the message is clear: technological progress must not inadvertently become a gateway for those seeking to circumvent the law. - FMT

P Sundramoorthy is a criminologist at the Centre for Policy Research at Universiti Sains Malaysia. He is an FMT reader.

The views expressed are those of the writer and do not necessarily reflect those of MMKtT.