Lessons from 2024’s Global Democratic Test

The headlines warned 2024 would be the year AI-generated disinformation destroyed democracy. With over 2 billion people voting across 50+ countries, experts predicted deepfakes would flood social media, synthetic audio would manipulate voters, and AI-powered influence operations would determine election outcomes.

Yet as 2024’s “super election year” concluded, the AI apocalypse had largely failed to materialise.

Analysis from Harvard’s Ash Center revealed that the dreaded “death of truth” did not happen – at least, not due to AI, while the Munich Security Conference noted that AI had a negligible impact and was used in mostly harmless ways.

But here’s the crucial insight from my Churchill Fellowship research: this relative calm shouldn’t breed complacency. My fieldwork across India, the UK, Ireland, and the United States during 2025 reveals that while AI didn’t deliver the dramatic disruptions predicted, it has fundamentally altered political discourse and trust in democracy in subtler yet significant ways.

Australia faces unique vulnerabilities. Unlike many democracies with periodic elections, we maintain a continuous electoral cycle: 20 elections scheduled across all nine jurisdictions from 2025-2028 alone. This creates persistent opportunities for malicious actors to test and refine AI-enabled influence operations, especially with electoral law fragmented across eight states and territories plus the federal level.

The question isn’t whether AI will impact Australian democracy, but how we prepare for the sophisticated threats emerging from this global testing ground.

My Churchill report will be published in the coming weeks – subscribe to my blog to get updates!