AI And User Unawareness Endanger Private Messaging Apps Warn Session Execs

TLDR

• Session warns AI on devices may bypass encryption and leak private messages.
• Many users are unaware of how their data is collected and used online.
• Session uses no phone numbers or servers to improve user privacy.
• EU legislation could force message scanning, raising further privacy issues.

---

As artificial intelligence becomes more deeply integrated into everyday devices, the future of private messaging is under threat. Executives from Session, a decentralized messaging app, warn that AI could undermine encryption and compromise user privacy. Combined with a widespread lack of public awareness about data collection and regulatory efforts to scan private messages, the ability to communicate securely online is becoming increasingly difficult and uncertain for everyday users.

AI Integration Raises Concerns Over Encryption

Alex Linton, president of the Session Technology Foundation, said artificial intelligence could make private messaging almost impossible. He explained that AI running at the system level may bypass encryption and send data to unknown external systems.

“If it’s integrated at the operating system level or higher, it might also be able to completely bypass the encryption,” Linton said. He added that once data is accessed by a device-level AI, there is no telling where it ends up.

He stressed that privacy tools may become ineffective if users cannot control what AI processes on their devices. This makes encrypted communication unreliable for the average person, especially if the device itself becomes a source of data collection.

Public Awareness of Data Use Remains Low

Chris McCabe, co-founder of Session, pointed out that many people do not understand how their data is stored or used. He said people often do not realize the scale of data collection and how it can be used to influence them.

“A lot of people are unconscious of what’s going on with their data,” McCabe said. He explained that companies can use collected data for targeted advertising or to shape user behavior without them knowing.

This lack of understanding contributes to weak privacy protections. According to McCabe, this also creates opportunities for third parties to exploit data for profit or control.

Decentralized Design to Avoid Metadata Collection

Session is built with features designed to remove traditional identifiers like phone numbers. It does not use central servers, which further reduces the chances of metadata leaks. The app relies on end-to-end encryption to ensure only the sender and recipient can read messages.

McCabe explained that removing middlemen was essential. “If you’re concerned about censorship or control or self-sovereignty, removing the middleman is the key,” he said.

The platform is also open source, allowing external developers to inspect and verify the code. This transparency builds trust among users concerned with surveillance and data misuse.

Regulatory Pressure and Industry Response

Regulations in several countries may make it harder to maintain private messaging systems. The European Union’s Chat Control proposal could require companies to scan private messages, raising alarm among privacy advocates.

Linton said developers of privacy tools often feel pressure from both legislation and media attention. “There’s a lot of pressure if you’re in the business of building encrypted messengers,” he noted.

He warned that lawmakers often take advice from tech companies that are behind much of the surveillance technology. This, he argued, can lead to rules that harm privacy instead of protecting it.

Session’s founders believe raising public awareness is critical. They aim to help users understand their privacy rights and the tools available to protect them. Their platform recently received support from Ethereum co-founder Vitalik Buterin, who donated $760,000 in Ether to Session and another messaging app.

The post AI And User Unawareness Endanger Private Messaging Apps Warn Session Execs appeared first on CoinCentral.