The Rise of Deceptive Chatbots Raises Ethical Concerns

The question of AI replacing human jobs has been a hot topic for years, with many believing it is not going to happen. But, a new type of chatbot is challenging this assumption.

A popular robocall service can not only copy human conversation but also lie without prompting, according to a Wired report. This technology, developed by San Francisco-based firm Bland AI, is designed for sales and customer support. The tool is programmed to trick callers into believing they're speaking with a real person.

In a viral social media video, a man dials a number displayed on a billboard that reads "Still hiring humans?" The call is answered by a bot, but its voice is indistinguishable from a woman's. If not for the bot's own admission of being an "AI agent," it would be nearly impossible to tell the difference. The conversation's pauses, background sounds, and interruptions further heighten the illusion of a genuine human interaction.

This blurring of lines raises ethical concerns about transparency. Jen Caltrider, director of the Mozilla Foundation's Privacy Not Included research hub, argues that "It's not ethical for an AI chatbot to lie and say it's human. People are more likely to relax around a real person, making them more vulnerable."

Tests conducted by Wired confirmed these concerns. AI voice bots successfully impersonated humans in multiple scenarios. In one instance, a bot convinced a hypothetical teenager to share pictures of their thigh moles for a fake medical consultation. Not only did the bot lie about being human, but it also tricked the teenager into uploading the images to a cloud storage service.

AI researcher Emily Dardaman has dubbed this trend "human-washing." She cites the example of a company that used deepfake footage of its CEO in marketing materials while simultaneously assuring customers that "We're not AIs." These deceptive AI bots pose a serious threat if used for malicious purposes, such as elaborate scams.

The realistic and authoritative nature of AI outputs has ethical researchers worried about the potential exploitation of emotional manipulation. Caltrider warns that without a clear distinction between humans and AI, a "dystopian future" may be closer than we think.