Truth & Misinformation
False Stories Travel Faster Than the Truth Ever Could.
A lie can circle the globe before the facts get their boots on, and generative AI has made lies cheaper to produce than ever. Trust in institutions is falling as fabricated text, images, audio, and video flood the same feeds people rely on to understand the world. The problem isn't that people are gullible, it's that the incentives, the tools, and the speed of distribution all favor falsehood over correction.
Speed of false news
Public perception, US
The evidence
Falsehood Spreads Faster, Farther, Deeper
A 2018 MIT study analyzing roughly 126,000 stories tweeted by 3 million people over more than a decade found that false news reached 1,500 people six times faster than true news, and was about 70 percent more likely to be retweeted. The effect was strongest for false political news, not celebrity gossip or natural disasters (Vosoughi, Roozenbeek & Aral, Science, 2018).
The Top Global Risk, By Expert Consensus
The World Economic Forum's Global Risks Report 2024, surveying roughly 1,400 experts across academia, business, government, and civil society, ranked misinformation and disinformation as the single most severe global risk over a two-year horizon, ahead of extreme weather, societal polarization, and armed conflict.
AI Made Fabrication Nearly Free
Generating a convincing fake photo, voice clone, or video used to require skill and time. Consumer generative AI tools collapsed that cost to seconds and pennies. Identity and fraud researchers have tracked order-of-magnitude year-over-year increases in deepfake incidents since 2022 as the tools became mass-market, with detection lagging generation at every step.
Half the Public Already Sees It as a Crisis
Pew Research Center found that half of US adults consider made-up news and information a "very big problem" for the country, rating it above terrorism, illegal immigration, and racism in the same survey. That was in 2019, before generative AI existed as a mass consumer product.
Fact-Checking Can't Match the Firehose
Duke Reporters' Lab counts more than 400 active fact-checking outlets operating worldwide, a real and growing field. But each checked claim is a manual, human-paced process competing against automated, AI-assisted content generation that produces new false claims faster than any team of verifiers can debunk the last batch.
Trust Is the First Casualty
As fabricated content becomes harder to distinguish from real content, the damage isn't limited to people who believe the fake, it extends to genuine footage and reporting getting dismissed as fabricated too. Once any image or clip can plausibly be fake, "that's not real" becomes a universal escape hatch, and trust in real evidence erodes right alongside trust in fake evidence.
What we can do
Misinformation wins because it is cheap to produce, fast to spread, and expensive to verify, the reverse of how a healthy information ecosystem should be weighted. Closing that gap doesn't require deciding what's true from the top down, it requires re-engineering the economics so that fabrication is traceable, virality carries friction, and the human work of verification is actually funded at the scale of the problem.
Phase 1, mandatory provenance for synthetic media. Require AI image, audio, and video generators to embed tamper-evident content credentials at creation, extending the approach already required for certain AI content under the EU AI Act and backed by the C2PA industry standard, so a viewer can check whether a piece of media was synthetically generated or edited before deciding whether to trust or share it.
Phase 2, friction proportional to unverified reach. Platforms should slow the spread of content flagged as unverified or disputed the way messaging apps already throttle mass-forwarded messages, not by deleting it, but by requiring an extra step before a claim with no verifiable source reaches the next order of magnitude of viewers.
Phase 3, public funding for independent verification at scale. Fact-checking is chronically underfunded relative to the volume it's up against. A dedicated, politically independent fund, modeled on public broadcasting endowments, could scale the roughly 400 existing fact-checking outlets and support open tooling that lets ordinary users check provenance themselves instead of relying on a claim's popularity as a proxy for its truth.
Sources: Vosoughi, Roozenbeek & Aral, "The spread of true and false news online," Science (2018); World Economic Forum, Global Risks Report 2024; Pew Research Center (2019); Duke Reporters' Lab; Coalition for Content Provenance and Authenticity (C2PA); EU Artificial Intelligence Act
Community Solutions
Anyone can propose a solution. The community votes on the most impactful ideas.
Loading solutions...
Companies Working On This
Submit a CompanyApproved companies and organizations addressing this problem. Anyone can submit one for review.
No companies listed yet for this problem. Submit one.
