Meredith Whittaker’s position on AI chatbots is unambiguous: they are tools, not confidants, and treating them as anything else carries real privacy costs. The Signal president made that argument in a Bloomberg interview published on 19 June 2026, conducted by Mishal Husain, in which she set out her objections to how the technology industry frames its AI products.
‘These are not your friends. These are not conscious beings. These are not sentient interlocutors,’ Whittaker said, responding to a question about the privacy implications of chatbots such as ChatGPT and Claude.
Why Meredith Whittaker Distrusts AI Chatbots for Real Thinking
Whittaker is not an absolutist. She acknowledged using AI tools ‘to format a document here and there.’ But she draws a firm line at outsourcing her reasoning to them: ‘I don’t ask them questions. I’m very serious about my thinking and writing, and I don’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 framing, a chatbot as a statistical average of prior output rather than a genuine intellectual interlocutor, runs counter to how many AI companies market their products. For Whittaker, the distinction is not merely philosophical. It touches directly on what she sees as Signal’s mission: preserving what she has called, in a Bloomberg weekend interview, ‘the fundamental human right to private communication’ in a world where ‘almost every micro-action is surveilled, turned into data, processed, modeled [and] sold.’
Copilot, Christmas Shopping, and the Backdoor Problem
The conversation turned pointed when Whittaker addressed a prediction by Mustafa Suleyman, Executive Vice President and CEO of Microsoft AI. Suleyman had suggested that users could let Microsoft Copilot handle all their Christmas shopping this year, with the assistant eavesdropping on family group chats to work out who wants what.
Whittaker was blunt about what that scenario actually requires. It means giving Copilot ‘access to my credit card, my browser, my Signal, the ability to message my siblings on my behalf, my home address [and] my calendar.’
‘What you’ve just described is a system with very pervasive access across multiple applications and services,’ she said. ‘In the context of Signal, it would constitute a kind of a backdoor.’
That word carries weight. Signal’s entire value proposition rests on the absence of such access. An AI agent that reads your encrypted messages to personalise a shopping list is, in Whittaker’s framing, functionally indistinguishable from a surveillance tool with a friendlier interface.
Suleyman’s ambitions for automation extend well beyond Christmas shopping. Fortune has reported his prediction that ‘white-collar work, where you’re sitting down at a computer, either being a lawyer or an accountant or a project manager or a marketing person, most of those tasks will be fully automated by an AI within the next 12 to 18 months.’ Whittaker’s critique of the Christmas-shopping scenario sits inside a much larger disagreement about how much autonomy AI systems should be granted over personal data and personal decisions.
Her concern about AI agents is not new. Speaking to Bloomberg at the World Economic Forum in Davos on 20 January 2026, Whittaker described deeper integration of AI agents into devices as ‘pretty perilous’ for encrypted services such as Signal, because such agents require access to large amounts of data stored across multiple apps. The Bloomberg interview published in June develops the same argument with sharper examples, as Suleyman’s Copilot vision gave her a concrete case to examine.
The exchange illustrates a genuine tension in how the technology industry and privacy advocates are approaching the AI agent moment. The industry’s pitch is convenience: let the assistant handle the friction. Whittaker’s counter is structural: the friction is not a bug. It is the barrier between your private life and a system designed to process and monetise it. How that argument lands with the hundreds of millions of people now using both Signal and AI tools will shape where the boundaries of encrypted communication end up being drawn.
