Renée DiResta: How to beat bad information

Episode Summary

Executive Summary: The episode examines how misinformation and disinformation spread online, especially around elections and COVID-19, and why social media’s speed, virality, and participatory culture make false narratives hard to contain. Rene DiResta explains how modern misinformation often emerges both top-down and bottom-up, can be amplified by ordinary users and pundits, and is increasingly complicated by deepfakes and AI-generated text. The discussion also highlights collaborative monitoring efforts to detect and contextualize harmful claims.

Main Topics: Misinformation vs. disinformation (Priority: 5/5): DiResta distinguishes accidental falsehoods (misinformation) from deliberate deception (disinformation), emphasizing that intent matters because it changes the source, motive, and response needed. Election falsehoods and legitimacy narratives (Priority: 5/5): The conversation focuses on post-2020 election claims, including narratives about voter fraud and the Sharpie-marker conspiracy, and how isolated irregularities were woven into large-scale claims of illegitimacy. Why social media accelerates falsehoods (Priority: 5/5): Platforms like Facebook, Twitter, TikTok, and Instagram provide tools for rapid creation, amplification, and community reinforcement, allowing narratives to spread faster than factual corrections. Bottom-up and top-down information flows (Priority: 4/5): The episode argues that propaganda is no longer only produced by institutions; grassroots communities can originate narratives that later get amplified by pundits, journalists, and mainstream media. Deepfakes and generative AI (Priority: 4/5): The discussion explores how synthetic video, audio, and text could intensify deception, while also noting current technical limits, detection methods, and the fact that deepfakes are often used for harassment and revenge porn. Media literacy and the next generation (Priority: 3/5): DiResta and Altman discuss whether younger users will become better at recognizing falsehoods, with media literacy emerging as a necessary skill as platforms and content formats proliferate. Multi-institution election monitoring (Priority: 4/5): The Stanford Internet Observatory’s coalition with academic, nonprofit, government, and journalistic partners illustrates a coordinated approach to identifying, contextualizing, and avoiding amplification of harmful narratives.

Key Arguments: Intent is crucial: false content shared innocently differs from deliberate deception used for profit, politics, or state interference. Election misinformation becomes especially dangerous when isolated incidents are connected into a conspiracy narrative that undermines trust in democratic transfer of power. Social media increases both the velocity and reach of misinformation, making corrections less effective and community reinforcement stronger. Modern misinformation is not purely top-down; grassroots groups can generate narratives that mainstream media then covers, creating a feedback loop. Pundit and influencer ecosystems act as a bridge between social media and mass media, often amplifying claims without journalistic verification. Deepfakes are concerning, but political deepfakes currently face technical imperfections and rapid forensic scrutiny; generative text may be a broader near-term risk. Media literacy is now essential because audiences must judge source credibility, partisan framing, and whether content is investigative or merely opinionated. Effective monitoring requires cross-sector collaboration among researchers, civil society, government, and journalists to contextualize claims and prevent unnecessary amplification.

Data Points: Year of election discussed: 2020 - The interview centers on misinformation and disinformation surrounding the 2020 U.S. election. Time since election: about a week and a half - DiResta describes the state of the post-election information environment at the time of the interview. CNN Facebook audience: about 35 million likes - Used to illustrate how traditional mass media also operate through social platforms. Youth platform example: TikTok - Highlighted as a venue where election misinformation appeared and spread among younger users. Number of core coalition partners: 4 - Stanford Internet Observatory partnered with University of Washington, Graphika, and DFRLab for election monitoring. Years since 2016: after 2016 - Government efforts like CISA and the FBI Foreign Influence Task Force were described as responses to 2016 election interference concerns. Age of DiResta's child: almost 7 years old - Used as an example of a generation growing up with cameras, YouTube, and normalized content creation. Election misinformation target scope: voting-related content - The monitoring coalition narrowly scoped its work to misinformation about voting process and legitimacy.

Pivotal Quotes: "Misinformation we mean information that's inadvertently wrong and then disinformation to mean information that's deliberately wrong." — Rene DiResta: Defines the central distinction used throughout the conversation. "What we have today is the phenomenon of the bottom up, where people in a Facebook group can come up with a theory or a claim... and so determined groups of people can make something hit the top by starting at the bottom." — Rene DiResta: Explains how online narratives can originate in grassroots communities and later spread upward into mainstream attention. "I'm an enthusiast. I'm not a journalist." — Russ Altman: Altman clarifies his role during the episode, underscoring the conversational rather than journalistic nature of the discussion.

Implications: Listeners should expect misinformation to remain a persistent feature of online life, especially around politics and health. The future response depends on media literacy, cross-platform monitoring, and better detection tools for synthetic media and AI-generated text.