Is Video Verification on Its Way Out?

When it comes to adopting new technology, today’s organizations constantly have to tread that delicate balance between security and convenience. Take video-based identity verification as an example. It has become a common practice in the financial services sector, including banks, lenders, and cryptocurrency providers, to utilize video calls for use cases like onboarding new customers to digital platforms or even providing high-risk authentication. For instance, this additional security measure is essential for high-risk transactions, such as cross-border transactions or those involving large sums of money, which carry a significant risk of financial loss.

On the one hand, this offers tremendous convenience for customers and employees, who can say goodbye to dealing with long lines and mountains of paperwork. On the other, recent strides in Generative AI (GenAI) technology have empowered cybercriminals to create deepfakes and exploit flaws in processes like video identification that were previously considered secure. While video call verification once enabled visual confirmation of user identities, seeing is no longer believing in the age of rampant GenAI content.

Deloitte, for one, cautions that deepfakes are being used to create copies of individuals’ voices or appearances so convincing that they can bypass remote authentication mechanisms, thus undermining their ability to ensure reliable identity verification. With a combination of ingenuity and a forged ID bolstered by the capabilities of GenAI, threat actors can superimpose artificial imagery on their faces to fool any verification technology that may lack the sophistication to detect such advanced techniques and the human operators overseeing it.

Deepfakes are also a hugely powerful tool in boosting the success of synthetic identity fraud, whereby fraudsters weave stolen credentials with manufactured information, often using AI to create fictitious personas. It’s a major concern for financial organizations, costing billions of dollars annually and often going undetected due to the lack of an identity owner, which is likely to be highly unreported.

What’s concerning is that deepfake technology is still relatively nascent. It will only get better — becoming more potent and convincing. And we know innovation cycles are faster than ever before. Therefore, the impact on organizations not equipped to defend against deepfake threats effectively can be significant, particularly within the financial services sector.

Early indications of this are already surfacing. For instance, in February, Arup, a global engineering company, lost over USD26 million when one of its Hong Kong-based finance employees fell victim to fraudsters who posed as his colleagues in a video conference call using deepfake technology. In December 2023, Singapore Senior Prime Minister Lee Hsien Loong and Prime Minister Lawrence Wong were targeted by deepfake videos, one promoting a cryptocurrency scam and the other endorsing an investment scam. As recently as June 2, Senior Prime Minister Lee Hsien Loong was again depicted promoting investment products in another deepfake scam.

These incidents only highlight how, in the wrong hands, technologies that power synthetic media can be used for nefarious ends that undermine trust. Hence, it isn’t surprising that there’s an increasing call to move away from video call identity verification methods entirely. But is that a viable solution, or merely avoiding the problem?

At the moment, GenAI is advancing at such a fast pace that there is now a constant stream of new tools entering the market. Many of these tools, such as OpenAI’s text-to-video tool, Sora, launched in February 2024, are making it much easier for threat actors to create high-quality video deepfakes that are becoming impossible to discern with the human eye.

Further contributing to the conundrum is the rising number of threat groups continuously working together to develop new, ingenious ways to outsmart remote identification systems. In a recent threat intelligence report, iProov analysts found that 47% of these groups were created in 2023 alone, indicating a surge of over 90% in the number of groups throughout that year.

It’s no longer a choice. Financial institutions must adopt more reliable verification and authentication techniques to protect themselves against these escalating threats. A great example would be to incorporate AI-based facial biometric matching and liveness detection, supplemented by human expert analyst oversight of the AI system. Combining AI-based technologies and expert human analyst oversight gives organizations a leg up on threat actors.

The encouraging news is that facial biometric verification technologies help organizations fight against deepfake-enabled attacks. For those on the lookout for such technologies, a range of solutions is available. Nevertheless, it’s vital to recognize that not all biometric face verification technologies are created equal. These technologies, like other security solutions, can provide a spectrum of identity assurance options and should be selected based on the risk of each use case.

In the effort to stay ahead of novel and constantly evolving threats, however, the key is to keep evolving and improving. This is best accomplished with SaaS-based solutions, as they can be updated as soon as new threat methodologies are discovered. Some SaaS vendors include active threat detection in response to the offerings, making them ideal for high-risk use cases.

As deepfake technology continues its exponential growth, organizations that rely on video authentication cannot afford to be complacent. Proactively implementing robust video verification measures is no longer a luxury but a necessity. By taking decisive action today, organizations can safeguard themselves from the devastating consequences of deepfake-induced fraud, misinformation, and reputational damage.

The views and opinions expressed in this article are those of the author and do not necessarily reflect those of CDOTrends. Image credit: iStockphoto/Mykyta Dolmatov