Why is Signal’s president critical of AI chatbots?

Meredith Whittaker believes chatbots lack consciousness and pose privacy risks when used for personal tasks. She fears data leaks and uncontrolled access to sensitive information.

What threats does Whittaker see in Microsoft Copilot?

Whittaker describes a scenario where Copilot gains access to credit cards, messages, calendars, and user addresses as a potential backdoor threat, risking confidential data exposure.

Does Signal’s president use AI tools?

Yes, but sparingly—only for tasks like document formatting. She avoids asking chatbots questions to prevent dependence on their responses for her own ideas.

← All news

Security

Signal Head Warns: AI Chatbots Are Not Your Friends

June 20, 2026

Photo: TechCrunch

Quick answer

Signal President Meredith Whittaker warns against trusting AI chatbots like ChatGPT and Microsoft Copilot due to risks of personal data leaks.

Signal President Meredith Whittaker expressed concerns about the growing popularity of AI chatbots like ChatGPT and Claude in an interview with Bloomberg. She argued that users often perceive these systems as friendly assistants, though they lack consciousness and meaningful dialogue capabilities. Whittaker emphasized she does not trust chatbots to shape her thoughts, preferring to work on texts independently.

She particularly criticized a scenario proposed by Microsoft AI head Mustafa Suleyman, where users could delegate holiday shopping entirely to Microsoft Copilot*. Whittaker noted this would require access to credit cards, browsers, Signal messages, calendars, and even the ability to send messages on a user’s behalf. Such integration, she warned, creates a backdoor risk for unauthorized access to confidential data.

Whittaker admitted using AI tools but only for routine tasks like document formatting. She stressed that serious thinking and analysis should not depend on chatbot responses, which merely average existing information. Her stance reflects growing cybersecurity concerns about AI’s widespread adoption in daily life.

Common questions

Why is Signal’s president critical of AI chatbots?: Meredith Whittaker believes chatbots lack consciousness and pose privacy risks when used for personal tasks. She fears data leaks and uncontrolled access to sensitive information.
What threats does Whittaker see in Microsoft Copilot?: Whittaker describes a scenario where Copilot gains access to credit cards, messages, calendars, and user addresses as a potential backdoor threat, risking confidential data exposure.
Does Signal’s president use AI tools?: Yes, but sparingly—only for tasks like document formatting. She avoids asking chatbots questions to prevent dependence on their responses for her own ideas.

Dzen feed: /feed/dzen.xml · RSS: /feed.xml