Global Roaming with Geraldine Doogue and Hamish Macdonald AI-generated content in political campaigns: how far will we let this go?
4 snips
Mar 30, 2026 Constanza Sanhueza, a political scientist at ANU who studies AI, misinformation and democratic backsliding, joins the conversation. She discusses how AI supercharges disinformation, the vulnerabilities of weaker democracies and youth on social media. She also explores foreign and domestic use of AI-driven political messaging and the challenges of building public digital literacy.
AI Snips
Chapters
Transcript
Episode notes
Investigator Traces Indonesian Disinformation To One Firm
- Ben Strick traced coordinated Indonesian accounts pushing West Papua messaging to a single marketing firm using reverse image searches and posting time patterns.
- He identified bot-like accounts (e.g., Marco267) posting identical content on a schedule and linked them to a company that owned many sites.
AI Amplifies Misinformation And Erodes Electoral Trust
- AI amplifies existing misinformation problems by enabling mass, fast distribution and realistic deepfakes that erode trust in institutions and elections.
- This reduces trust both directly (people believing false content) and indirectly (believing others are misled), undermining confidence in outcomes.
Weaker Democracies Suffer More From AI Disruption
- AI's negative effects on information ecosystems hit weaker democracies and autocracies harder, though established democracies feel impact during election cycles.
- The Varieties of Democracy data show censorship and media bias worsened where AI-enabled tools spread disinformation.