The two conferences differed markedly in scale and scope. MisinfoCon, sponsored by a consortium of nonprofits, held 10 sessions during the week, each one discussing a specific initiative or group of similar initiatives related to the fight against misinformation and disinformation. (Attendance wasn’t available at press time.) MozFest was organized by the Mozilla Foundation, which also produces and maintains the Firefox browser through its for-profit Mozilla Corp. subsidiary. It attracted 7,300 attendees to 340 sessions on nine thematic tracks, or “Spaces.” In keeping with Mozilla’s mission to “ensure the Internet is a global public resource, open and accessible to all,” these nine Spaces covered a range of topics, including Rethinking Power & Ethics and Sustainability & Climate Change. One of the nine was Misinformation & Disinformation, which offered 18 sessions from Monday through Thursday. Taken together, MisinfoCon and the MozFest Misinformation & Disinformation Space delivered a fire hose of good news and bad news from the current battle lines of the war on information.

The Supply Side

One way to combat misinformation and disinformation is to choke off the supply of it: identifying it, correcting it, and/or having platform operators remove it. This is a hard problem, because the purveyors of misinformation and disinformation, like drug dealers, scammers, and other bad actors, are constantly developing new strategies and tactics. The battle can degenerate into a game of whack-a-mole—a fact referred to more than once during the week. Still, there were reports on robust supply-side interdiction efforts and some successes.

As the internet’s premier source of facts, Wikipedia figured prominently in several sessions. Miriam Redi and Diego Saez-Trumper, researchers with the Wikimedia Foundation, discussed the challenges of identifying knowledge gaps and protecting knowledge integrity in Wikipedia. They also reported on the deployment of AI to help enforce Wikipedia’s rule that potentially controversial statements must be backed by a reliable source.

In a separate session, a panel of Wikipedia volunteer editors described their experiences. One, a specialist in Nazi World War II history, offered a vivid account of the persistence of misinformation and disinformation in her field. She noted that the presence of references doesn’t guarantee accuracy, giving an example of a seemingly valid reference to support a controversial statement—except that the reference didn’t support the assertion. She also described being harassed by another editor whose misinformation she repeatedly corrected. Ultimately, the other editor was banned from Wikipedia—for harassment, not for misinformation. Continuing the discussion of references, another editor offered her practice of avoiding the problem of dead links by requesting that the Wayback Machine preserve pages being referenced. Another editor who specializes in health and medicine mentioned that it’s much easier to deal with bad information than missing information. She tries hard to identify and fill gaps, because misinformation and disinformation capitalizes on them.

Aside from fact-checking, another way to throttle disinformation is to choke off its funding. One panel on this topic described the estimated $400 billion digital advertising industry and how ad networks and brokers enable the ads of mainstream companies to appear on sites that spread misinformation, without the companies realizing it. The panel included the co-founders of the Check My Ads Institute, which uncovers and combats the tactics being used to fund disinformation.

Several programs featured discussions of the differences among social media platforms. Kate Starbird of the University of Washington described the Election Integrity Partnership, which monitored the response of major platforms to complaints about misinformation during the 2020 U.S. presidential election. Of the four platforms monitored, Facebook had the worst response rate, at 76%. Another panel discussed the unique problems of misinformation on TikTok. Because of its unique functionality, the panelists found TikTok to be a very effective tool for amplifying misinformation and extremist content. The TikTok problem has been compounded by the refusal of the company’s leaders to acknowledge the problem. They maintain that the platform isn’t a source for political content and therefore hasn’t established effective procedures for addressing problematic content.

The global dimensions of the misinformation and disinformation problem were highlighted in multiple sessions. One presented updates from fact-checking and public awareness initiatives in Nigeria and Kenya, along with Viral Facts Africa, a pan-African initiative sponsored by the World Health Organization. These efforts, which have focused on combating health misinformation and disinformation during the pandemic, have subsidized internet connectivity and in some cases paid social media influencers to promote accurate information. Another session featured panelists from the Global Voices Civic Media Observatory. Its network of local editors aims to provide “local contextual knowledge, subtext and language analysis to emerging events and narratives” and “[identify] harmful information, elevating the good.” One of the organization’s projects has been to compile a database of misinformation promoted by the government of Brazil.

The fight against misinformation and disinformation has attracted for-profit intelligence analysis and consulting firms as well as nonprofits. At least three were identified in the program. John Kelly, CEO of Graphika, presented examples of the sophisticated analysis and visualization work done by his firm, using mostly open source data. He also referred to training resources available from another for-profit firm, Bellingcat. In another session, analysts from ActiveFence discussed their large-scale monitoring of disinformation. They characterized the diversification of actors and media, noting that disinformation campaigns have spread beyond the botnets of a few years ago. One of the tools available to these campaigns is the Random Face Generator, a resource for generating photos to accompany fake personas.

The Demand Side

Besides attacking the supply of misinformation and disinformation, the other way to thwart it is on the demand side, by reducing people’s appetite for it. This is being done primarily through education in information literacy or digital literacy. The demand side was the subject of only two sessions. One, an international panel of IFLA leaders, encouraged audience ideas and experiences via breakout discussions and collaborative editing of a Google Doc on challenges and approaches in media literacy education. The other, led by librarians from Rowan University, described their initiatives in embedding the teaching of lateral reading in the curriculum. Their work is based on a Stanford University research and demonstration project, which has been documented in various sources.

Conclusion

The bad news from MisinfoCon and MozFest is that the worldwide war on information is continuing to intensify and will be with us for a long time. The distribution of misinformation and disinformation—deliberate or mistaken, malevolent or innocent—is a worldwide problem. It involves a proliferating cast of characters, operating on a diversifying set of platforms, using ever-evolving tactics. Nobody has solved this problem yet. But the good news is that the defenders of information are robust too. The problem is much better understood than it was a few years ago, and there are more initiatives to address it. Countermeasures are evolving. Judging by the presentations, most of the effort is going into choking off the supply of misinformation and disinformation, but there’s promising news on the demand side too. Let’s hope that future conferences highlight continued progress.