Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Signal president warns the hyped agentic AI bots threaten user privacy

Mar 08, 2025 - businessinsider.com
Meredith Whittaker, president of Signal, expressed concerns about the security risks posed by agentic AI, which refers to AI agents capable of performing tasks and making decisions without human input. Speaking at the SXSW 2025 Conference, Whittaker highlighted the privacy implications of allowing such AI systems access to vast amounts of user data, as they would require permissions to access browsers, credit card information, calendars, and messaging apps like Signal. She warned that these AI agents could potentially undermine user privacy by processing data off-device and accessing multiple databases without encryption.

Whittaker's concerns are echoed by Yoshua Bengio, a prominent AI researcher, who also warned about the risks associated with agentic AI at the World Economic Forum in Davos. Bengio emphasized the need for scientific understanding and technological investment to ensure the safe development of AI, cautioning against the potential dangers of creating systems that could surpass human intelligence and pose existential threats. Both experts stress the importance of addressing these privacy and security challenges before advancing further with agentic AI technologies.

Key takeaways:

  • Meredith Whittaker, president of Signal, warns that agentic AI poses significant security and privacy risks due to its need for extensive data access.
  • Agentic AI refers to autonomous bots that can perform tasks and make decisions without human input, potentially accessing sensitive user data.
  • Whittaker highlights the potential for agentic AI to undermine privacy by requiring root-level access to various personal data sources and possibly processing data off-device.
  • Yoshua Bengio, a prominent AI researcher, echoes concerns about agentic AI, emphasizing the need for scientific understanding and technological investment to mitigate risks.
View Full Article

Comments (0)

Be the first to comment!