Rohan Phadte (left) and Ash Bhat (right)Courtesy of Robhat Labs

In a world where the most famous dorm-room-born internet company has developed a reputation as a matrix of fake users and misleading posts, Ash Bhat and Rohan Phadte are hoping that the answer to online disinformation could come out of their own college apartment.

Bhat and Phadte, both 21, are the founders of Robhat Labs, which they launched while previously students at the UC Berkeley. Last year, they debuted two misinformation-fighting projects. The first is NewsBot, an app for Facebook Messenger that aims to identify the political leaning of a given news piece. The second is Botcheck.me, a Chrome extension that uses machine learning and natural language processing to rate whether Twitter accounts are likely to be fabricated bots. In less than a year, Botcheck.me has flagged nearly 1 million accounts as probable bots.

The duo’s third project, set to be released next month, is a free browser extension called SurfSafe. As Bhat explained on Thursday at the Aspen Ideas Festival, which is co-hosted by the Aspen Institute and The Atlantic, SurfSafe uses the images on a news article as a “signature” to find similar pages from around the web. It then uses textual analysis to comment on how the article compares to other coverage, and based on that, whether it might be trustworthy. SurfSafe will also help users to identify whether an image was pirated from another site and flag visuals that have been photoshopped.

SurfSafe will allow its users to identify websites they find reliable as their own “source of truth,” Bhat said on Thursday. He argued that, even if a user indicates that she only trusts unreliable sites, she’ll start to notice after a while that they can conflict with each other, and she’ll become suspicious that they aren’t trustworthy. After a while, he claimed, “people will start to move closer to the objective truth.”

When I asked how Robhat Labs would react if its users defy its creators’ expectations and start rating biased sources as trustworthy, and respected, fact-checked new sites as compromised, Phadte was even-keeled. “If the people are making those judgments, there must be some reasoning behind it,” he said. If that happens, Phadte said, the company would be “updating our models accordingly.” In an effort to make SurfSafe as “completely objective as possible,” the team is hoping to rely on algorithms and AI. As Phadte put it, in all of Robhat Labs’ endeavors, “we don’t want to introduce a lot of human-created bias.”

Although Bhat and Phadte were optimistic about tech companies’ efforts to fight misinformation on their sites, they acknowledged that “the incentives are definitely misaligned.” Despite pledges by Mark Zuckerberg and other company heads to do better, fake accounts and fake stories continue to propagate online. “We have to reengineer these kinds of tools to make sure that we are not misleading the public,” Phadte said—and it was much easier to understand “we” to be a batch of hopeful 20-somethings than engineers at Facebook or Twitter.

But those hopeful 20-somethings might be reluctant to get into what Bhat called the “fake-news market,” especially since, as Bhat said, the space is largely viewed as unprofitable. “We don’t have very many competitors right now,” Bhat said, “and we hope it does change.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.