The rapidly increasing prevalence of AI-generated content and deepfakes has left many questioning everything they see online. In fact, as much as 72 per cent of consumers worry on a day-to-day basis about being fooled by a deepfake into handing over sensitive information or money.

New data from Jumio, the automated, AI-driven identity verification and compliance solutions provider, comes from its new 2024 Online Identity Study, the third instalment of its annual global consumer research.

Jumio results this year highlight significant concerns among consumers about the risks associated with generative AI and deepfakes, including the potential for increased cybercrime and identity fraud.

Jumio surveyed over 8,000 adult consumers, split evenly across the United Kingdom, United States, Singapore and Mexico. The study found that only 15 per cent of consumers said they’ve never encountered a deepfake video, audio or image before, while 60 per cent have encountered a deepfake within the past year.

Even as people become increasingly concerned about deepfakes, 60 per cent of consumers still believe they could detect a deepfake, up from 52 per cent the year before. This suggests that a significant proportion of consumers are overestimating their ability to recognise what is a deepfake and what is not.

Men emerged as more confident in their ability to spot a deepfake (66 per cent of men versus 55 per cent of women), with men aged 18-34 demonstrating the most confidence (75 per cent), while women aged 35-54 were least confident (52 per cent).

Sixty per cent of consumers want more governmental regulation of AI to address these issues. However, regulatory trust varies globally, with 69 per cent of Singaporeans expressing trust in their government’s ability to regulate AI, compared to just 26 per cent in the UK, 31 per cent in the US and 44 per cent in Mexico.

The impact of fraud

Stuart Wells, CTO at Jumio, said: “As generative AI advances, the incidence of deepfakes continues to rise, revealing a significant gap in our collective ability to detect these deceptions. This continued overconfidence underscores the critical need for stronger public education and more effective technological solutions. It’s essential that businesses and consumers collaborate to enhance digital security measures to effectively prevent identity fraud.”

Sixty per cent of consumers want more governmental regulation of AI to address these issues. However, regulatory trust varies globally, with 69 per cent of Singaporeans expressing trust in their government’s ability to regulate AI, compared to just 26 per cent in the UK, 31 per cent in the US and 44 per cent in Mexico.

Overall, 68 per cent of respondents either know or suspect that they’ve been a victim of online fraud or identity theft, or that they know someone who has been affected. US consumers were most likely to be direct victims of fraud (39 per cent) either knowingly or by suspicion, and Singapore was the top country to report knowing a victim (51 per cent).

While 46 per cent of the consumers who were or suspected they were a victim of online fraud or identity theft said the ordeal was a minor inconvenience, 32 per cent said it caused significant problems and several hours of administrative work to resolve, and 14 per cent went as far as calling it a traumatic experience.

Regardless of whether they’ve been a victim of fraud or identity theft, most consumers worry daily about falling victim to data breaches (79 per cent) and account takeover attacks (77 per cent).

Finding a solution

Identity verification is a key part of the solution for companies looking to secure themselves and ensure that their users are genuine. Over 70 per cent of consumers said they’d spend more time on identity verification if those measures improved security in industries including financial services (77 per cent), healthcare (74 per cent), government (72 per cent), retail and e-commerce (72 per cent), social media (71 per cent), the sharing economy (71 per cent), and travel and hospitality (71 per cent).

When creating a new online account, global consumers said taking a picture of their ID and a live selfie would be the most accurate form of identity verification (21 per cent), with creating a secure password coming in at a close second (19 per cent).

Philipp Pointner, chief of digital identity at Jumio, said: “As we navigate the complexities introduced by generative AI, the role of sophisticated security systems becomes crucial. To counter the rise in deepfakes and cyber deception, incorporating multimodal, biometric-based verification systems is imperative. These technologies are key to ensuring that businesses can protect their platforms and their customers from emerging online threats, and are significantly stronger than passwords and other traditional, outdated methods of identification and authentication.”

The power of artificial intelligence (AI) is undeniable, but in the wrong hands, it can be devastating. To combat this possibility, Jumio, the automated, end-to-end identity verification, risk assessment and compliance solutions provider, has launched its Jumio 360° Fraud Analytics at Money20/20 USA.

According to Jumio’s analysis, 25 per cent of fraud is interconnected — either perpetrated by fraud rings or by individuals using the same information or credentials to open new accounts. These can be on banking sites, e-commerce platforms, sharing economy sites and more. To help combat this, the new Jumio feature uses AI-driven predictive analytics to identify fraud patterns with more sophistication and accuracy.

“Today we are pushing the industry to evolve once again by harnessing the power of predictive analytics to identify fraud patterns across our vast network. This will allow us to identify and stop fraud rings and other coordinated attacks with more accuracy than ever before.” – Robert Prigge

How it works

It looks beyond simple linkages, such as flagging someone because they are connected to a known fraudster. Instead, it looks at billions of data points across Jumio’s cross-industry network to identify patterns based on behavioural similarities and other indicators.

Using graph database technology with a layer of machine learning, it groups identity transactions into clusters across the network and determines the fraud risk of each cluster. This provides a multi-dimensional view of each transaction and the cross-customer ecosystem as a whole.

Further features include:

• AI-driven predictive analytics determine the likelihood of fraud risk for each identity transaction based on its cluster association
• The predictive fraud score can automatically reject identity transactions. Furthermore, it can stop fraudsters from accessing your platform in real time
• Dashboards provide transparency and help you visualise connected data

A worthy addition

Today, when a user goes through the ID verification process, Jumio performs a wide variety of advanced fraud checks. With the addition of Jumio 360° Fraud Analytics, the identity transaction will also be compared to the clusters and generate a predictive fraud score that can be used to automatically reject the transaction if it exceeds a certain threshold.

Jumio’s initial studies show that this approach improves its existing, highly accurate fraud detection rate by at least 30 per cent — without increasing the false rejection rate. Business users can audit the reasons behind the decision in the Jumio Portal or via APIs.

Jumio 360° Fraud Analytics is currently available in early release to select customers. It will be generally available in early 2024.

“Because explainability is a key requirement when using automation, we provide insights on why the identity transaction was rejected,” said Philipp Pointner, Jumio chief of digital identity. “We also provide a graphical representation of the high-level linkages we found. This makes it easy to see connections between people, documents, devices and more. The combination of sophisticated analytics and easily consumable insights makes this new technology a game-changer for fighting fraud.”

“Jumio revolutionised online identity verification when it pioneered the ID + selfie approach,” said Robert Prigge, Jumio CEO. “Today we are pushing the industry to evolve once again by harnessing the power of predictive analytics to identify fraud patterns across our vast network. This will allow us to identify and stop fraud rings and other coordinated attacks with more accuracy than ever before.”