馃敀 COP-GPT? OpenAI might report users to the police
OpenAI has acknowledged that conversations held in ChatGPT are monitored and in some cases may be shared with the police. The information, published in a blog post, sparked a wave of outrage among users and experts.
As OpenAI reports, content submitted by ChatGPT users that is deemed dangerous is sent to human reviewers who can decide to notify law enforcement. As the company emphasizes in its blog post, the procedure only applies to “direct threats to life”. However, it’s unclear how the location of users is determined or what specific content might trigger such an intervention.
Cited by futurism.com, critics point out that this practice is at odds with previous statements by OpenAI CEO Sam Altman, who promised “therapist-level privacy”. Doubts also arise about the risk of abuses, like impersonating others to trigger false interventions.
Share
You might be interested in
-
馃敀 A calculator for minds or systemic laziness? The essence of the debate about AI in schools
While Asian schools are turning education over to algorithms en masse, we are still seeking a humanistic compromise. The skills of the entire next generation of workers depend on which model we…
-
馃敀 AI in creative work: muzzle or overdrive?
In the debate over tools for generating text, images or music, two extreme fears collide. There’s growing fear of fake content, copyright infringements and the loss of artists’ control over their own…
-
馃敀 Regulatory straitjacket or digital shield? Two visions for the future of AI
The world is facing a dilemma: whether to keep AI in check with regulations, or to trust free-market mechanisms and legal liability. For businesses, this means calculating whether it is more cost-effective…
-
馃敀 AI’s green tax: Will data centers face special taxation?
The issue of AI鈥檚 impact on the climate has ceased to be the domain of activists and has made it onto CFOs’ agendas. Is AI another industrial revolution that requires freedom, or…
