This week, a German tabloid tested the boundaries of AI passing itself off as reality. Celebrity magazine Die Aktuelle published a cover story in its April 15th issue about a supposed interview with Formula One driver Michael Schumacher; only at the end does it reveal that it was a hoax produced entirely by an AI chatbot. Schumacher’s family told ESPN it plans to take legal action against the gossip rag.
The article promised “the first interview” with Schumacher, who suffered a severe brain injury on a family ski trip in the French Alps in 2013. Since then, the celebrated driver — arguably the greatest in Formula One history — hasn’t appeared publicly as his family guards his privacy.
The public’s best description of Schumacher’s life today came from the 2021 Netflix documentary Schumacher, in which his wife Corinna Schumacher said, “Michael is here. Different, but he’s here, and that gives us strength, I find.” She continued, “We live together at home. We do therapy. We do everything we can to make Michael better and to make sure he’s comfortable, and to simply make him feel our family, our bond.” Concluding, she stated, “We’re trying to carry on as a family, the way Michael liked it and still does. And we are getting on with our lives. ‘Private is private,’ as he always said.”
This week’s issue of Die Aktuelle, showcasing a photo of a smiling Schumacher, promised “No meagre, nebulous half-sentences from friends. But answers from him! By Michael Schumacher, 54!” according to ESPN. A subheading on the magazine cover reads, “Es Klingt täuschend echt,” which translates to “It sounds deceptively real.”
Tabloids testing the waters of AI-generated bunk is likely only the tip of the iceberg. As generative AI rapidly advances, separating fact from fiction — whether through chatbot-generated text, fake photos or (very soon) convincing bogus videos — will become increasingly difficult.
Without clear legal and regulatory boundaries, the world’s fascists, propagandists and other nefarious actors will exploit generative AI mercilessly for money and power. For example, what happens if fake images, deep-fake videos or audio recordings about political candidates go viral as voters head to the polls in the pivotal 2024 US election? Without any meaningful consequences for abuse, we find ourselves in a new AI Wild West. And although the Biden administration appears to be taking some initial steps in exploring safeguards against this type of behavior, I’m skeptical that America’s senior government leadership will skate to where the puck will be, rather than falling hopelessly behind this quickly evolving new form of misinformation and disinformation.