Episode 105: Truth vs. Disinformation in the Age of Generative AI
April 24, 2025 · 1 min read
The rise use of generative AI raises significant challenges around disinformation. In this episode, we'll explore the differences between misinformation, disinformation, and malinformation. How can detection tools, digital signatures, provenance tracking, and education help combat these issues? How can emerging technologies like self-sovereign identity, Web3 and open data shape a more secure and informed online world?
#Speakers
- Host: Mike Schwartz, Founder/CEO Gluu
- Guest: Harrison Tang, Co-Founder / CEO Spokeo
- Guest: Joe Andrieu, Board of Directors Rebooting the Web of Trust
#Homework
- Opinion Disinformation 2.0 in the Age of AI: A Cybersecurity Perspective--Why disinformation is a cyber threat. pdf
- Navigating Truth in a Post-GenAI World with Self-Sovereign Identity
#Takeaways
- Misinformation: false information spread by people who think it's true; Disinformation: false information spread by people who know its false for malicious intent; Malinformation: true information spread with malicious intent (for example publishing an activist's home address)
- Trust is built over time through repeated interactions and consistent behavior, to creat a mutually beneficial outcome (i.e. a "win-win")
- While LLMs have defeated liveness detection (the best AI detection tools are only 95% effective), they can't defeat algorithmic protections like cryptography (or solve chess problems).
- The benefit of open standards for wallets is the potential for credential diversity. Joe pointed out that the Web was better then AOL and Compuserve because of the unmatchable amount of content. Perhaps with open standards, we'll see a profusion of new credentials to solve problems like disinformation.