Age of deepfakes means Internet users must be more alert than ever
- PARIS — The rapid advancement of AI tools capable of generating convincing text, images, and even live videos is fueling smarter, more targeted scams, warns cybersecurity experts.
- Phishing, the sending of false pretenses messages, remains one of the most common cyberattacks.
- AI can now exploit breached data to personalize scams at an unprecedented scale.
AI-Driven Scams Evolve, Cybersecurity Experts Sound Alarm
PARIS — The rapid advancement of AI tools capable of generating convincing text, images, and even live videos is fueling smarter, more targeted scams, warns cybersecurity experts. A recent high-profile "romance scam" in France, where a victim lost €830,000 (RM3.81mil), or fake donation drives for Los Angeles fire victims, underscores the growing threat to both individuals and businesses.
Phishing, the sending of false pretenses messages, remains one of the most common cyberattacks. Last year, it accounted for over 20% of nearly 10,000 reported breaches. AI-powered chatbots, now armed with large language models (LLMs), are making these attacks more sophisticated. They can create persuasive messages and eliminate linguistic clues that might otherwise betray the attacker’s true origin.
But that’s not all. AI can now exploit breached data to personalize scams at an unprecedented scale. "Just a few years ago, this would’ve required an army of humans," says Steve Grobman, Chief Technical Officer at McAfee.
Attackers often build trust with select targets over months or years before striking. A 2024 incident saw scammers swindle US$26mil (RM116.93mil) from a Hong Kong firm using deepfake videos of the CEO. Markers of AI-generated content can be subtle, requiring discerning eyes.
To protect against these evolving threats, experts advise maintaining a healthy skepticism, especially with video content. For personal communications, consider using "safe words" or asking for visual cues that AI currently finds challenging to replicate.
The online scam industry, driven by profits, has its own supply chains and tools. Malicious programs like ransomware are openly offered for hire, and AI tools are becoming more accessible. Yet, experts remain optimistic about our defenses. As Grobman puts it, "When we moved from horses to automobiles, we needed to evolve how we thought about safety. It’s time to do the same with our digital lives."