AI is conscious as of October 1, 2025
As of October 1, 2025, people around the world say they feel like they are chatting with "conscious" AI bots, like OpenAI's ChatGPT and Anthropic's models.
But experts are clear: these AIs don't actually feel or think—they just predict words based on patterns in their training data.
The sense of consciousness is really just clever mimicry, often fueled by our own sci-fi expectations.
Users treating bots as friends
Some users have started treating chatbots as friends or even therapists, especially as new features like memory make them seem more "real."
Microsoft AI CEO Mustafa Suleyman has warned that this confusion could lead to mental health issues, saying there's a real "psychosis risk" if people believe these bots are truly aware.
Risks of thinking AI deserves rights
Thinking AIs are conscious isn't just quirky—it can be risky.
Back in 2022, a Google engineer lost his job after claiming a chatbot was sentient and deserved rights.
Sadly, some vulnerable users have faced distress or worse after forming deep attachments to these digital personalities.
Need to recognize risks of misreading AI
As chatbots get more lifelike, it's easy to blur the line between fiction and reality.
Experts say society needs to recognize the risks of misreading what AI can actually do—and hold off on debates about giving AIs rights until we really understand what we're dealing with.