Signal President Meredith Whittaker calls out agentic AI as having ‘profound’ security and privacy issues

Signal President Meredith Whitaker warned that Friday Agent AI The user may be at risk of privacy.

Speaking at the stage at the SXSW conference in Austin, Texas, advocate for protected communication, referred to the use of AI agents as “keeping your brain in a jar” and warned that this new example of computing – where AI performs work on behalf of users – both privacy and protection are “deep problems”.

Whiter explained how AI agents are being marketed as a way to add prices to your life by conducting various online tasks for the user. For example, AI agents will be able to look for concerts, booking tickets, determining the event on your calendar, and the tasks that your friends have booked.

“So we can only put our brain in a jar because the thing is doing it and we don’t have to touch it, isn’t it?” Whiteker is difficult.

He then explained that the type of access to the AI ​​agent would have to perform these tasks with our web browser and perform these tasks, as well as to send a text to your friends, including access to our credit card information for our calendar and messaging app.

“It needs to be able to drive [process] Whitaker warned.

“And if we’re talking about strong enough … the AI ​​model that is strengthening it, but no way is happening on the device,” he said more. “It is almost certainly being shipped to the cloud server where it is being processed and returned. Thus there is a deep problem with protection and privacy that is suffering this hype around the agents and it is threatening to break the blood-brain barrier between applications and OS layers in the end [and] Cutting their data, “Whitaker reached the conclusion.

He said if a messaging application like a signal was integrated with AI agents, it would ease the privacy of your messages, he said. The agent will have to access the app to send your friends text and to summarize the data to summarize those texts.

He commented on how the AI ​​industry was built on the surveillance model with mass data collection in his comments earlier during the panel. He said that “bigger AI Paradism” – which means more data, better – the possible consequences that he did not seem to be better.

Whitecker warned that “We want to reduce more privacy and protection in the name of” magic Jenny Bot that is about to take care of the excuses of life, “he said in conclusion.