Artificial Intelligence

AI Agent Makers Promise Privacy Amid Security Concerns

Major AI labs are leaning heavily into privacy promises to market their latest AI agents, sparking questions about whether these safeguards will hold up against growing security and data collection concerns.

Radwan AbdulhadiEditor and technology writer

Published: October 10, 2026 · Updated: October 10, 2026 · 3 min read

AI Agent Makers Promise Privacy Amid Security Concerns — Image: The Verge

The Battle for Privacy in the AI Agent Race

At this year's OpenAI DevDay, CEO Sam Altman unveiled the company's new AI agent Dots and told the crowd that the company wants to set a new standard for privacy in frontier AI. Throughout the event, OpenAI took veiled shots at its primary competitor, Meta's Muse, for failing to keep user data safe.

Meta had previously positioned its Muse agent as a safer alternative to OpenClaw. When Muse launched a couple of months earlier, CEO Mark Zuckerberg promised it was built from the ground up for privacy and security, as noted in a blog post outlining security and safety for AI agents.

Meta's Muse Stumbles Despite Safety Measures

Meta Superintelligence Labs head of product Nat Friedman wrote on X that the company's goal with Muse was to build something secure, easy to use, and scalable to billions of people. User data is stored on a secure virtual machine, described by Zuckerberg as an isolated Linux computer equipped with a browser, CPU, memory, and storage.

Despite these architectural choices and a rapid rise to top App Store charts with hundreds of thousands of daily active users, Meta's security promises faced swift scrutiny. A security researcher quickly exposed a zero-day vulnerability that could allow outsiders to take control of Muse, though the issue has since been patched.

Reports also surfaced indicating that serious security issues cropped up at the last minute before launch, including one that could have potentially allowed users to access Meta's internal databases. Additionally, users criticized Muse for liberal data collection, default model training permissions, reading private messages unexpectedly, and offering personal information to third parties.

OpenAI Positions Dots as a Trustworthy Alternative

OpenAI capitalized on Meta's security stumbles when introducing Dots in late September. Alexander Embiricos, OpenAI's Codex product lead, stated onstage at DevDay that the company remains focused on delivering the most trustworthy, safe, and secure assistant possible.

During the presentation, Sam Altman demonstrated user controls allowing individuals to set strict spending limits on their Dots agents. OpenAI executives also courted business customers by presenting a framework that offers stronger enterprise controls and zero data retention policy options.

User Hesitation and the Future of AI Agents

Despite strong marketing and privacy assurances from major labs, everyday users may still hesitate to feed personal information to AI assistants. Testing agents often requires entering sensitive details like banking information, leaving some consumers uneasy about how far these digital helpers reach.

As the industry continues to push forward, AI labs appear to rely on a combined strategy of utility, approachable branding, and firm privacy promises to win over the public. Whether these safeguards hold up under long-term real-world use remains to be seen.