UK Universities Face ‘deepfake’ Threat as AI Automates Interviews

Automated Applications Cut Costs but Face Rising ‘Future of Fraud’ Threat

UK Universities Battle ‘Deepfake’ Threat in AI-Powered Admissions

As UK universities increasingly rely on automated online application processes to streamline admissions for international students, a new and alarming challenge has emerged—deepfake fraud. Advanced artificial intelligence (AI) technologies are enabling applicants to manipulate online interviews, posing a serious threat to the integrity of the admissions process.

The Rise of Deepfake Applicants

In an effort to save time and money, many UK universities have turned to automation for screening international applicants. Platforms like Enroly, which are widely used to facilitate this process, have detected cases where AI-generated deepfake technology was used to alter applicant responses in pre-recorded video interviews.

Enroly recently reported that a small number of candidates employed deepfake techniques to superimpose AI-generated faces and voices over real individuals. This method allows applicants to enhance their fluency, disguise their identities, or even have someone else answer questions on their behalf. Such fraud, while still in its infancy, is a concerning development for universities and immigration authorities alike.

The Existential Crisis of UK Universities

The Scale of the Problem

Although deepfake fraud remains relatively rare, its potential impact is significant. Enroly uncovered about 30 cases out of 20,000 interviews conducted for the January 2025 intake. While this represents only 0.15% of applications, the company warns that it is evidence of a growing trend that could escalate in the future.

According to Phoebe O’Donnell, Enroly’s head of services, deepfake technology poses a substantial challenge for university interview assessors. “It’s the stuff of nightmares—fake faces layered over real ones, complete with expressions and movements,” she explained. “It’s like something out of a spy film. And yes, they’re incredibly hard to detect. But hard isn’t impossible. Thanks to real-time tech and a few clever tricks up our sleeves, we’ve already stopped several attempts.”

Traditional Fraud vs. Deepfake Deception

While deepfakes are an emerging form of fraud, more conventional tactics remain prevalent. Enroly’s analysts found that outright deception occurred in 1.3% of the 20,000 interviews. Common fraudulent practices included:

• Impersonation – Applicants having someone else take their interview.
• Lip-Syncing – Using pre-recorded audio while mimicking speech.
• Third-Party Support – Receiving off-camera assistance during interviews.

Despite the small number of deepfake cases, the increasing sophistication of AI-generated fraud signals the need for stronger detection measures.

How Universities Are Fighting Back

To combat fraudulent applications, UK universities are implementing advanced verification technologies, including:

• Facial Recognition Technology – Matching applicant faces with official identification documents.
• Real-Time Monitoring – Using AI-driven tools to detect unusual facial movements or inconsistencies in real-time.
• Live Interview Follow-Ups – Flagging suspicious pre-recorded interviews for manual review and live interviews.
• AI Pattern Analysis – Identifying behavioral anomalies that may indicate deepfake usage.

These tools are crucial in maintaining the credibility of the UK’s higher education system and ensuring compliance with UK Visas and Immigration (UKVI) regulations.

The Stakes for UK Universities

The use of automated interviews is not just about efficiency—it’s also about compliance. UK universities risk severe consequences if fraudulent applications go undetected. Under UKVI regulations, if more than 10% of a university’s sponsored applicants are refused a visa, the institution could lose its ability to issue Confirmation of Acceptance for Studies (CAS) certificates, a requirement for student visa approval.

Automated interviews have become a key part of the pre-CAS screening process, allowing universities to assess applicants’ English proficiency and overall credibility before issuing a CAS. By identifying potential fraud at an early stage, universities can avoid visa refusals that could jeopardize their sponsorship status.

The Future of Fraud in University Admissions

The rise of deepfake technology is a reflection of broader concerns about AI-driven fraud. Last year, the CEO of WPP, the world’s largest advertising group, fell victim to a deepfake scam involving an AI-generated voice clone and a fake video representation of him on a conference call. If high-profile business leaders can be deceived, universities must remain vigilant against similar risks in student admissions.

As deepfake technology evolves, fraudsters will continue to refine their methods, making detection more challenging. Universities, software providers, and regulatory bodies must work collaboratively to stay ahead of these advancements. Regular updates to fraud detection algorithms, increased training for admissions staff, and multi-layered verification processes will be essential in mitigating risks.

Final Thoughts

While AI-powered automation has revolutionized university admissions, it has also introduced new vulnerabilities. Deepfake fraud, though currently limited in scope, represents a potential threat to the credibility of UK higher education institutions. By leveraging cutting-edge detection tools and reinforcing manual review processes, universities can protect themselves from this emerging form of deception. What’s next? Universities and tech providers must remain proactive in addressing AI-driven fraud, ensuring that automation enhances efficiency without compromising integrity. As the landscape of cyber deception continues to evolve, staying ahead of fraudsters will be an ongoing challenge in the digital admissions era.