When hackers use AI: IAA Trendtalk shows human firewall as the key to security
The 62nd General Assembly of the International Advertising Association (IAA) in Zurich focused on the digital threat landscape. Criminal analyst Mark T. Hofmann provided a compelling demonstration of how hackers use AI and deepfakes—and m&k conducted an exclusive interview with him.
Zurich - The International Advertising Association (IAA) sent out a clear signal at its 62nd General Assembly: Artificial intelligence is not only a driver of innovation, but also a tool of cybercrime. Renowned crime analyst and business psychologist Mark T. Hofmann made it clear in his presentation that humans remain the biggest weak point. In an exclusive interview with m&k, Hofmann also shared the most important recommendations for the communications industry.
New board members elected
Alice Sauerer, Head of Marketing Campaigns at Swisscom, and Tom Hagendorn, Senior Manager at Team Farner, were newly elected to the Board of Directors at the Annual General Meeting. President Matthias Kiess (CEO TBWA) welcomed the new additions and emphasized their fresh impetus. Sauerer wants to contribute her experience as an advertising client, while Hagendorn wants to drive forward innovative programs for young talent in the industry. The existing members were confirmed unanimously. With the departure of Michel Siegenthaler (Swisscom) and Philipp Wild (Webrepublic), the board was reorganized, with Wild remaining a member of the association.
The dark side of AI
The focus of the event was the presentation by Mark T. Hofmann: "The dark side of AI: How hackers use AI & deepfakes". His analysis showed how the quality and quantity of attacks by AI is increasing. Cybercrime has long operated like a multi-billion dollar industry with organizational structures, financial departments and even "customer service". According to forecasts, the global damage caused by cybercrime could grow to 10.5 trillion dollars by 2025.
According to Hofmann, around 90 percent of all attacks are based on human error. Hackers don't "crack" systems, they "crack people". Typical attacks follow a four-ingredient formula: They pose as a trusted authority, create emotion and time pressure and link this to an unusual request.
Deepfakes as a game changer
Hofmann paid particular attention to the risks posed by deepfakes. Today, 15 to 30 seconds of audio material is enough to clone voices deceptively realistically. This not only makes CEO fraud easier - cases such as a 35 million dollar bank fraud in Dubai are already a reality - but also political disinformation and new forms of traditional fraud are possible. Even video recordings could lose their evidential value in court in the future.
Prevention and the "human firewall"
Hofmann called for cyber security to be seen as a top priority. In addition to technical standards such as long passwords, two-factor authentication and regular updates, awareness is needed above all. Cyber drills such as a fire alarm, prepared emergency plans, alternative communication channels and physical protective measures are essential. However, it is crucial to design awareness in such a way that it also reaches those who are not interested in IT security.
Exclusive m&k interview
In an interview with m&k, Hofmann emphasized that companies should not rely on technology alone in the future: "The human firewall is the strongest protective shield. It is crucial that employees recognize the patterns of attacks - and dare to question them." Hofmann explains the specific steps that managers and communication teams should take in detail in the interview (see video/post).
Conclusion
The IAA Trendtalk 2025 combined political innovations with a deep insight into the mechanisms of modern cyber attacks. The event underlined the role of the IAA as a platform for orientation - and at the same time showed that digital resilience remains first and foremost a question of human vigilance.
