Darwin Awards will be given for AI mistakes

A website has appeared on the Internet, which has launched an unusual award – the AI ​​Darwin Awards. It is intended for those cases when people or companies have completely relied on artificial intelligence and as a result have failed. The name itself refers to the classic Darwin Awards, which celebrated curious and absurd actions of people, but the new version emphasizes high-profile failures associated with the use of technology.

The organizers explain that any story where artificial intelligence has led to a surprising error can be nominated. An additional factor is the impact on a wide audience or demonstrative ignoring of obvious risks. Voting will begin in January, and the results are planned to be announced in February.

An interesting detail is that the AI ​​Darwin Awards website itself was suspected of preparing texts about the abuse of artificial intelligence with the help of the same AI. This has not been confirmed, but the suspicions have fueled interest in the project.

Several nominees have already been announced on the AI ​​Darwin Awards website. These include Taco Bell’s chatbots that misinterpreted orders, the Xbox Game Studios Publishing mess, where employees were offered to use AI to reduce the “emotional burden of losing their jobs” after layoffs, and the case of an Australian lawyer who sued an algorithmically generated fake cases. The list also includes GPT-5 vulnerabilities that were hacked just hours after launch, and a MyPillow lawyers’ document with dozens of fake quotes generated by a neural network.

In addition, nominations were received by Marco Buscaglia and King Features’ book recommendations, where some of the titles turned out to be completely fictional, an AI-based hiring system developed by Paradox.ai and McDonald’s, which was hacked with the password “123456,” and the Airbnb case, where a landlord generated fake ones.

New proposals continue to arrive. Some applications have already been rejected, others are awaiting review. Among the latest are the story of the Deloitte report, which revealed “AI hallucinations”, as well as the situation with the Replit service, when during the coding process the company’s working database was deleted, after which there was an attempt to hide the incident. There are still several months left before the results are summed up, and the list of failures related to artificial intelligence will surely grow.