In this multi-part research project, Dr. Sarah Kreps and I studied how synthetic (AI-generated) disinformation can deceive the public and masquerade as reliable, human-written news.
- Wrote “All The News That’s Fit to Fabricate: A Study of Synthetic Disinformation” with Dr. Kreps and Miles Brundage (published in Cambridge’s Journal of Experimental Political Science).
- Co-authored “Not Your Father’s Bots” in Foreign Affairs with Dr. Kreps.
- Co-authored “Taking GPT-2 head-to-head with the New York Times” in Brookings’ TechStream with Dr. Kreps.
- Contributed to OpenAI’s official GPT-2 release report, “Release Strategies and the Social Impacts of Language Models”. (Dr. Kreps and I are listed as authors; we wrote the section on human perception and detection.)
- Contributed to (and cited by) OpenAI’s 6-month follow up to GPT-2.