Vietnam faces a significantly higher rate of deepfake-related cyberattacks than the global average, highlighting growing pressure on businesses to strengthen cybersecurity capabilities as artificial intelligence (AI) becomes increasingly integrated into cybercrime.

A global cybersecurity survey released by security company Kaspersky on November 19 found that AI-powered cyberattacks are becoming more prevalent across the Asia-Pacific (APAC) region.

Traditional phishing remained the most common attack method in the region, accounting for 24% of reported incidents.

More sophisticated techniques, including attacks exploiting AI vulnerabilities and deepfake technology that impersonates individuals through manipulated images and voices, each accounted for 14%.

Vietnam recorded a particularly high proportion of deepfake-related incidents at 21%, nearly double the global average of 11% and considerably above the regional figure of 14%.

Human error remains Vietnam's biggest cybersecurity weakness

The survey also revealed significant vulnerabilities related to employee cybersecurity awareness.

In Vietnam, 75% of surveyed organisations reported experiencing attacks that exploited weaknesses stemming from employees' lack of security awareness.

This was the highest proportion in the Asia-Pacific region, exceeding the regional average of 63% and the global average of 60%.

The findings suggest that businesses face growing exposure to increasingly sophisticated fraud techniques, particularly those involving AI-generated impersonation.

In response, Vietnamese organisations are demonstrating strong interest in AI-powered cybersecurity solutions.

According to Kaspersky, 87% of organisations in Vietnam expressed openness and willingness to invest in security tools incorporating AI, the highest rate in the region.

Malaysia followed at 76%, while Indonesia recorded 74%.

However, Kaspersky warned that the technical capabilities of cybersecurity teams are not developing as quickly as investment in new technologies.

As a result, companies may struggle to fully utilise advanced security tools or maximise their effectiveness against emerging threats.

Adrian Hia, Managing Director for Asia Pacific at Kaspersky, emphasised that AI should support rather than replace cybersecurity professionals.

"AI can help professionals maximise their capabilities, but it cannot replace human judgement, awareness and practical understanding," he said.

He described these abilities as essential skills that IT teams across the region urgently need to strengthen.

AI tools cannot replace human judgement

Cybersecurity experts noted that AI systems rely on previously trained data, which can limit their effectiveness against unfamiliar attack techniques or newly developed impersonation methods.

Automated systems also cannot fully replace human expertise in critically evaluating security alerts, assessing the business consequences of data breaches or explaining cybersecurity decisions to senior management.

These limitations are changing how companies recruit and train cybersecurity professionals.

Across the Asia-Pacific region, practical training and professional experience are now the two most highly valued qualifications among employers, each cited by 76% of respondents.

Professional certifications followed at 74%, while university degrees were valued by 57%.

The survey also found that 70.4% of information security professionals consider soft skills equally important as technical expertise.

The findings reflect a growing recognition that cybersecurity professionals need more than technical knowledge to identify threats, respond effectively to incidents and communicate risks within organisations.

Cybersecurity workforce faces critical skills shortages

Businesses across the region continue to experience substantial shortages of cybersecurity specialists with essential technical capabilities.

According to the survey, the most significant gaps involve system programming skills, reported by 62% of organisations, followed by knowledge of software development frameworks at 61% and networking technologies at 46%.

Soft skills are also an area of concern.

Analytical thinking was identified as a weakness by 48% of respondents, while 43% highlighted shortcomings in problem-solving abilities.

These deficiencies could limit organisations' capacity to detect, investigate and respond to increasingly complex cyberattacks.

To address the growing threat of AI-powered fraud and deepfake attacks, experts recommended moving away from standardised cybersecurity training programmes that apply the same approach to every employee.

Instead, businesses should establish specialised professional development pathways tailored to different cybersecurity functions.

These include security operations, threat intelligence, malware analysis and network security.

Role-specific training would help cybersecurity teams develop the expertise needed to operate AI-powered tools effectively while strengthening their ability to identify sophisticated impersonation attempts and other emerging threats.

As cybercriminals increasingly exploit AI technologies, Kaspersky's findings underscore the importance of combining advanced security systems with skilled professionals and stronger employee awareness.

Du Lam