Penn Center on Media, Technology and Democracy published research by Dan Roth digging into why asking an LLM the same question two different ways is a cheap, effective way to catch when it doesn't actually know the answer.
Public source
Publisher name
Public post
Is Google AI killing the open internet? Nilay Patel coined the term “Google Zero”—when web traffic to publishers plummets as people stay on search-embedded chatbots. Dan…
Company
Penn Center on Media, Technology and Democracy
We use cutting-edge science to understand the information ecosystem and leverage that research to strengthen democracy.
- Industry
- Technology, Information and Internet
- Location
- Philadelphia, US
- Company size
- 2–10 employees
About Penn Center on Media, Technology and Democracy
The Penn Center on Media, Technology and Democracy is committed to independent research on the information ecosystem and its impact on democracy. Our mission is to advance our collective understanding of the information landscape through cutting-edge science—then leverage that research to foster a more informed society and strengthen the foundations of democracy. We focus on the media organizations and the online platforms, the incentives and the algorithms, that determine which ideas and beliefs spread, and which don’t. The Center’s faculty and scholars are distinguished by their use of emerging technology for data-driven research of the information ecosystem, including novel data collections, large-scale cloud computing, and innovative applications of AI.
See moreLatest activity
Latest activity from Penn Center on Media, Technology and Democracy
5 signals
Research & Knowledge
Penn Center on Media, Technology and Democracy published research by Matt Levendusky and Kathleen Hall Jamieson revealing that doubting the legitimacy of one election can harden into a stable belief that colors how people perceive subsequent elections.
Research & Knowledge
Penn Center on Media, Technology and Democracy published research by Danaé Metaxa and Ro E. finding that LLMs are not reliable sources of abortion information, with accuracy weakest in abortion-restrictive states.
Research & Knowledge
Penn Center on Media, Technology and Democracy published research by Chris Callison-Burch and Delip Kumar showing that search-enabled LLMs still get roughly half of academic citations wrong, leaning on internal memory over the sources they actually retrieve.
Discover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
ISD (Institute for Strategic Dialogue)
ISD (Institute for Strategic Dialogue) researchers tested leading AI models—including Gemini, Grok and GPT-5.5—on 2,400 prompts tailored to 10 states, assessing accuracy, timeliness, ability to debunk contested claims, use of authoritative sources, and performance when questions were asked in Spanish rather than English.
Research & Knowledge
Advancing Humans with AI
Advancing Humans with AI published a study from the MIT Media Lab following 67 people over four weeks as they judged news headlines and images with an AI chatbot.
Research & Knowledge
Lenz
Lenz.io conducted research where five frontier AI models fact-checked 1,000 real-world claims and provided self-ratings on a 1–10 scale.
Research & Knowledge
Siensmetrica, Inc.
Siensmetrica, Inc. published a multi-year study in Nature Scientific Reports revealing a critical vulnerability in generative AI under sustained, multi-turn conversational pressure.
Research & Knowledge
Brennan Center for Justice