Signal President Meredith Whittaker is pushing back against the growing tendency to treat artificial intelligence chatbots as companions, confidants, or conscious partners. Her message is direct: the systems millions of people are opening up to are none of those things.
What Whittaker Is Actually Saying
Whittaker's warning centers on three specific denials. AI chatbots are not friends. They are not conscious beings. They are not sentient interlocutors — meaning they are not genuinely understanding, experiencing, or responding the way a living entity would.
That three-part framing matters because it closes off the most common emotional assumptions people bring to these tools. The "friend" framing implies loyalty and mutual care. The "conscious being" framing implies awareness and inner experience. The "sentient interlocutor" framing implies genuine comprehension and reciprocity. Whittaker, speaking from her position at Signal, is saying none of those assumptions hold.
Why This Warning Carries Weight
Signal is an encrypted messaging platform whose mission centers on privacy and user protection. Whittaker's institutional position gives her warning a structural character that distinguishes it from ordinary tech skepticism: this is not a warning from someone with a financial stake in AI products, but from an organization whose model depends on users understanding what their communication tools actually are — and are not.
Companies building AI chatbots have strong commercial incentives to encourage emotional engagement. More engagement means more usage. Whittaker's position sits outside that incentive structure, which makes her skepticism different in kind from the enthusiasm generated by the industry itself.
The Practical Stakes for Users
When users mistake a chatbot for a friend or a conscious being, they may share information they would otherwise guard, make decisions based on the assumption of genuine understanding, or develop emotional dependencies on systems that are neither aware of them nor capable of the reciprocity they imagine.
The distinction Whittaker is drawing is not philosophical hairsplitting. It has real consequences for how people handle their own data, their emotional well-being, and their expectations of what these tools can and cannot deliver. Naming what AI chatbots are not, she is arguing, is a precondition for using them clearly.