Conan O’Brien joins the fight against corporate AI scam attacks

Somewhere in your inbox, there’s a tedious cybersecurity training that you clicked through during your onboarding process and have likely forgotten about. To combat scam attacks and boring cybersecurity training, cybersecurity AI firm Adaptive Security partnered with comedian and Emmy-winning late-night host Conan O’Brien on a 15-video training series meant to help employees recognize and stop scam attacks. Filmed in Los Angeles and co-written by O’Brien’s media network Team Coco, the videos rely on the comedian’s trademark humor to keep watchers engaged on an alarming topic. “We try to make that training a lot more engaging, a lot more entertaining, to actually drive results rather than it just be something that someone sits through because it’s mandatory,” says Andrew Jones, Adaptive’s cofounder and chief product officer. Founded in 2024 by Jones and Brian Long, Adaptive works with more than 1,000 companies and raised more than 140 million from investors including startup investment funds from OpenAI and Nvidia, Bain Capital Ventures and Andreessen Horowitz. Adaptive hopes that O’Brien’s training modules will “meet the employee where they are” by helping them retain information about a threat that is only increasing with advances in AI. “Voice phishing, deepfake personas and AI-generated impersonation are hitting companies every week, and most training programs were built years before those attacks existed,” Long, Adaptive’s CEO, said in a press release. “We wanted to build something personal and engaging, something employees would actually look forward to.” As AI models get better and cheaper, Jones added that many companies are concerned that the attacks are going to get more sophisticated. “These attacks are increasingly more sophisticated than they were even six months ago or a year ago,” he told Fast Company. “When we talk to our customers, they’re all very worried about these attacks—particularly the attacks that are powered by AI.” These worries are not unfounded. In 2024, the FBI’s Internet Crime Complaint Center (IC3) reported more than 16 billion in cybercrime losses. And according to a 2024 Deloitte report, U.S. fraud losses could more than triple from 12.3 billion in 2023 to 40 billion in 2027 because of generative AI. Phishing or email spoofing and invoice fraud are just a few of the most well-known types of corporate cyber scams. In the post-AI world, these scams have grown more advanced, especially with the advent of deepfakes, which Jones calls “the most scary.” Deepfakes allow scammers to use audio or visual impersonation tactics that feel very realistic—like a phone or Zoom call or a Microsoft Teams chat. “AI also allows the attackers to automate these attacks so they can run 24 hours a day, seven days a week,” Jones added. AI also allows for scams to be increasingly multimodal and personalized. Not only might employees receive scams in different formats, but it also allows an attacker to collect private information on someone, like where they live or what school they went to. AI deepfakes or pervasive on social media and in the entertainment industry, too. Bad actors can clone a public figure’s voice or likeness to disseminate disinformation, spread non-consensual explicit images and even AI endorsement fraud for brands or fake giveaways. In the ‘90s, one of O’Brien’s most popular comedy bits involved him “deepfaking” celebrities by replacing their mouths in still photos with his. “I creeped out an entire nation,” O’Brien said in Adaptive’s intro video. “Back then, making a deepfake took hours. But today, AI can clone a person’s voice and face in seconds.” “I teamed up with Adaptive Security just to figure out what these kids are up to,” O’Brien said in a press release. “Turns out it’s pretty cool.” Adaptive’s technology uses AI models to power its products and enterprise services including training modules, phishing simulations and risk scoring. “The nature of this problem is that it impacts every business and every vertical,” Jones said. “The commonality amongst all the conversations is [that] they’re seeing an increase. They’re very worried about the sophistication of the attacks, and they want to make sure that their employees are prepared.”