AI-Powered Deepfakes Are Fooling CEOs on Zoom Calls. 2026 Is the Most Dangerous Year Online.
North Korean hackers used AI deepfakes of a CEO on a live Zoom call. Phishing kits cost less than Netflix. Former Interpol director says AI has supercharged crime.
The AI Post newsroom — delivering AI news at the speed of intelligence.
In early 2026, Google's cybersecurity researchers spotted a group linked to North Korea using an AI-generated deepfake of a prominent CEO on a live Zoom call to fool a victim into compromising their computer security. The deepfake was not a pre-recorded video. It was a real-time interactive participant in a video conference. Two years ago, this was impossible. In 2026, it is routine.
This is not an isolated incident. It is part of what security experts are calling the fifth wave of cybercrime, driven almost entirely by AI tools that have turned once-rare attack capabilities into commodity products available to anyone.
The Numbers Are Staggering
A January report from cybersecurity firm Group-IB found that criminals can now acquire phishing kits on the dark web for the price of a Netflix subscription. These "synthetic identity kits" include AI video actors, cloned voices, and even biometric datasets. Former Interpol Director of Cybercrime Craig Jones warned that AI has "dramatically increased the speed, scale, and sophistication" with which criminals operate in 2026.
The attack surface has expanded in every direction. AI-powered social engineering is creating hyper-personalized impersonation attacks that mimic targets' friends, family members, and colleagues with unprecedented accuracy. Deepfake CEOs on video calls. Cloned voices requesting wire transfers. Phishing emails so personalized they reference last week's actual meetings.
Our Take
The AI industry has spent years talking about hypothetical existential risks from superintelligence. Meanwhile, the actual existential risk for millions of people is already here: AI-powered crime that makes the internet genuinely unsafe. You cannot trust a phone call from your bank. You cannot trust a Zoom call with your CEO. You cannot trust an email from your colleague. Every channel of human communication is now compromised.
The companies building these AI models need to own this problem. They built the tools that made deepfake CEO calls possible for the price of a Netflix subscription. Talking about "responsible AI" while criminals use your technology to steal billions is not a strategy. It is complicity through negligence.
Sources: The Independent, Google Cloud Threat Intelligence, Group-IB.