Across Asia, cybercriminals deploy increasingly advanced deepfakes to execute high-stakes fraud. Governments and businesses have responded by investing heavily in defensive artificial intelligence (AI) technologies. This escalating arms race has transformed the region into a testing ground for next-generation security tools that are reshaping how organizations protect themselves from threats.
The new face of corporate fraud in Asia
When deepfake videos of Singapore’s Prime Minister Lee Hsien Loong began circulating online, they looked disturbingly real. The fabricated clips showed him endorsing investment schemes that didn’t exist. Thousands of viewers saw the videos before authorities could intervene, part of a coordinated campaign designed to exploit trust in one of Asia’s most digitally advanced economies.
The deception reached new levels in Hong Kong. A finance worker joined what appeared to be a routine videoconference with colleagues to discuss a confidential transaction. The familiar faces on screen looked right. The voices sounded right. Only after transferring $25.6 million across 15 transactions did the employee discover the truth — everyone else on the call had been a deepfake. Criminals combined voice cloning and video manipulation to create a flawless illusion, bypassing authentication protocols that businesses had trusted for decades.
The scale of the problem continues to grow. Deepfake scam volumes have increased by over 2,000 percent in the past three years, according to Ashwini Karandikar, a principal at Oliver Wyman who tracks financial fraud across Asia. However, volume tells only part of the story. The real danger lies in adversarial AI, where criminals reverse-engineer the very tools designed to stop them.
Investigators testing security vulnerabilities took an image of a panda and changed a few pixel values. The alterations were invisible to the human eye, but when they fed the image into a facial recognition platform, the neural network confidently identified it as a gibbon. Financial institutions rely on similar technology to authenticate customers and flag suspicious activity. If criminals can manipulate those inputs the same way, legitimate users might be locked out while fraudsters sail through.
The threat has also become industrialized. AI-powered cyberattacks now come from organized cartels that function like tech startups, complete with research and development cycles. They test attack methods, analyze what succeeds, and refine their approach. Techniques and tools circulate across borders through networks that mirror legitimate open-source communities. Each breakthrough accelerates the next.
A multibillion-dollar market takes shape
Commercial opportunities emerge from the deepfake threat. The deepfake AI market in Asia-Pacific is projected to reach $6,139.7 million by 2033, driven mainly by demand for detection and prevention technologies.
According to Interpol’s cyberthreat assessment for the Asia and South Pacific region, rapid connectivity expansion, combined with uneven cybersecurity maturity, enables transnational criminals to exploit infrastructure gaps. However, many countries racing to digitize their economies lack the security capabilities to match, creating demand for solutions that can scale quickly across diverse technical environments.
Venture capital firms and corporate investors are directing substantial resources toward startups that develop authentication technologies. Security specialists and established technology companies alike pursue this opportunity by expanding their product portfolios. Banks, telecommunications providers, and government contractors compete to deliver similar or even better solutions.
A legal and technological counteroffensive emerges
Governments and private-sector innovators are responding with regulations and technologies that drive security development through clear standards and funding.
Governments enact stricter laws
South Korea passed some of the most aggressive deepfake laws in 2024. Legislation in that year expanded criminal liability beyond creation and distribution to include viewing or possessing certain nonconsensual sexual deepfakes.
China took a different regulatory approach focused on disclosure and platform accountability. The government banned nonconsensual deepfakes entirely and now requires clear identification when synthetic media is used for legitimate purposes.
Authorities took action against applications like Zao when they failed to meet identification requirements. After privacy concerns emerged, the face-swapping app was blocked on WeChat and faced regulatory scrutiny.
New technologies rise to meet the threat
Detection algorithms are one of the most active areas of development in the fight against deepfake threats. By analyzing inconsistencies in lighting, facial movements, and audio synchronization, these tools can flag manipulated content before it causes damage. Machine learning models train on vast datasets of authentic and synthetic media to identify telltale artifacts.
Digital watermarking creates another line of defense. Researchers have created methods that embed invisible markers in authentic content during creation. These techniques maintain reliable detection capabilities. The embedded watermarks demonstrate resilience against adversarial attempts to remove or distort the hidden information. Now, content creators have a way to prove authenticity even after images or videos circulate online.
Teams also rely on biometric authentication. Instead of using solely what appears on screen, such tools require users to complete live challenges that deepfake technology cannot keep up with, at least for the moment. A good example is the multifactor biometric checks banks implement for high-value transaction workflows.
The road ahead for innovation in AI security
Asia’s deepfake arms race is unlikely to slow down anytime soon. Both attackers and defenders will continue to refine their capabilities in an escalating competition that will influence the region’s AI sector for years to come. Organizations that invest in robust defenses stand to benefit as the security market expands.
Zac Amos is the Features Editor at ReHack, where he covers business tech, HR, and cybersecurity. He is also a regular contributor at AllBusiness, TalentCulture, and VentureBeat. For more of his work, follow him on X (Twitter) or LinkedIn.
TNGlobal INSIDER publishes contributions relevant to entrepreneurship and innovation. You may submit your own original or published contributions subject to editorial discretion.
Featured image: Mihail-Anton Ghiga on Unsplash