OpenAI’s Dots makes a privacy promise

At this year’s OpenAI DevDay, CEO Sam Altman unveiled Dots, the company’s new AI agent, and said OpenAI wants to “set a new standard for privacy in frontier AI.” The announcement put privacy at the center of the product’s pitch, as competition among AI agents increasingly turns on promises to keep users’ data safe.

OpenAI spent the day taking veiled shots at Meta’s Muse, its primary competitor, for failing to protect users’ data. But the privacy claims in this race are not one-sided: Muse had launched a couple of months earlier as a supposedly safer alternative to OpenClaw, with Meta CEO Mark Zuckerberg promising it was “built from the ground up for privacy and security.”

That puts Dots’ promise in a broader context. OpenAI is presenting its agent as part of a push for a higher privacy standard, while challenging a rival that had made a similar pledge about its own product. The announcements establish competing claims, rather than showing whether either company can deliver on them.

OpenAI takes aim at Meta’s Muse

At OpenAI’s DevDay, CEO Sam Altman unveiled the company’s AI agent, Dots, and said OpenAI wanted to “set a new standard for privacy in frontier AI.” The announcement also came with pointed criticism of Meta’s Muse, Dots’ primary competitor. OpenAI argued that Muse had failed to keep users’ data safe, casting its own product as the more privacy-conscious option.

The contrast is notable: OpenAI’s pitch for Dots directly challenges Muse’s privacy claims, rather than simply introducing another AI agent. The company positioned Dots as a safer alternative to Muse, while using the launch to question whether Meta had delivered on protecting users’ information.

But the competing assurances are part of a broader pattern. Muse had itself been introduced as a safer alternative to OpenClaw, with Meta CEO Mark Zuckerberg saying it was “built from the ground up for privacy and security.” Now OpenAI is making a similar promise about Dots while criticizing Muse’s handling of user data. The claims put privacy at the center of the rivalry between the agents.

Muse was pitched as safer than OpenClaw

OpenAI’s privacy pitch for Dots lands amid a chain of competing claims about AI agents. A couple of months before Dots was unveiled, Meta launched Muse as a supposedly safer alternative to its predecessor, OpenClaw. Meta CEO Mark Zuckerberg said Muse was “built from the ground up for privacy and security.”

That framing matters because Muse is now the product OpenAI is indirectly challenging. At DevDay, Sam Altman said OpenAI wanted to “set a new standard for privacy in frontier AI,” while the company took veiled shots at Muse over whether users’ data is safe. The contrast is striking: Muse had itself been introduced with a promise of stronger privacy and security, and Dots is now being positioned against it with a similar assurance.

The sequence leaves a broader question hanging over the claims. If one agent is presented as a safer successor to another, and a competitor then promises to raise the privacy standard, what should users look for beyond the assurances? The article sets up that tension without resolving it: companies are making privacy central to their pitches, but the competing promises still need to be judged.

The question behind competing privacy claims

The competing pitches raise a central question: can AI agent makers deliver on their privacy promises at a time when companies hoard customers’ personal data? OpenAI says Dots is intended to set a new privacy standard, while criticizing Meta’s Muse for failing to keep users’ data safe. But Muse arrived with its own assurance: Mark Zuckerberg said it was built from the ground up for privacy and security.

That claim came as Meta positioned Muse as a safer alternative to OpenClaw. Now OpenAI is presenting Dots as a safer alternative to Muse, turning privacy into a point of competition between the companies. The sequence shows how assurances can be layered: one product is promoted as safer than its predecessor, then challenged by a rival making a similar promise about its own agent.

The question is not only which company makes the strongest claim, but whether users’ data will actually be protected. The article’s competing promises leave that unresolved: declarations of privacy and security are easy to make, while delivering on them is the test facing AI agent makers.