Meredith Whittaker does not care if you think your AI chatbot is your friend. She wants you to know it’s not.
In a recent interview with Bloomberg, the Signal president drew a clear line in the sand about what chatbots like ChatGPT and Claude actually are, and what they definitely are not. “These are not your friends. These are not conscious beings. These are not sentient interlocutors,” she said flatly.
That’s the kind of blunt take you expect from someone running one of the most privacy-focused messaging apps in the world. But what makes Whittaker’s stance interesting is that she’s not some technophobe clutching her flip phone. She actually uses AI tools, just not the way most people do.
“I don’t ask them questions,” Whittaker explained. She’ll use them to format a document here and there, sure. But when it comes to thinking through ideas or doing actual writing? No thanks. She’s “very serious” about her thinking and doesn’t want the process of working through an idea to be “foreclosed or eclipsed by the response of a system that’s averaging what’s already out there.”
That’s a sharp observation, and it’s worth sitting with for a moment. When you ask an AI to help you think through something, you’re essentially crowdsourcing your own thought process with a mirror of everything that’s already been said on the internet. Whitter’s point isn’t that the technology is useless, it’s that there’s a real cost to outsourcing your reasoning to something that can only average what’s already there.
This skepticism extends far beyond her own usage. Whittaker had some pointed thoughts about Microsoft AI CEO Mustafa Suleyman’s recent prediction that users could let Microsoft Copilot handle all their Christmas shopping this year. You know, the scenario where Copilot is quietly eavesdropping on the family group chat to figure out who wants what.
The problem, as Whittaker sees it, is what that kind of access actually entails. Giving Copilot the ability to handle your Christmas shopping means giving it “access to my credit card, my browser, my Signal, the ability to message my siblings on my behalf, my home address and my calendar.” That’s not a helpful assistant. That’s a system with “very pervasive access across multiple applications and services.”
And in the context of Signal, she didn’t mince words: this would constitute “a kind of a backdoor.”
It’s a striking framing. Signal built its reputation on end-to-end encryption and minimal data collection, precisely because it recognizes that giving any system that level of access is a fundamental privacy risk. Microsoft might frame Copilot as a convenience, but Whittaker sees it as something much more consequential.
What makes this worth paying attention to is that Whittaker isn’t some distant critic. She operates inside the Technology industry and understands exactly how these systems work and where the pressure points are. Her perspective cuts through the breathless hype around AI assistants as helpful companions and reminds us that every shiny new feature comes with tradeoffs we should actually think about.
The bigger question no one seems eager to answer is what happens when these AI systems become indispensable and the data they’re collecting becomes too valuable to refuse. Convenience has a way of eroding privacy one small acceptance at a time.
