Editorial illustration for Social media platforms struggle with AI election lie containment
FEC Website Down as AI Election Lies Spread
Social media platforms struggle with AI election lie containment
The Federal Election Commission's website went dark for parts of this week after hackers disrupted it, according to the latest edition of MIT Technology Review's The Download newsletter. The outage lands at an awkward moment for an agency already stretched thin trying to track AI-generated falsehoods spreading across Facebook, X, and TikTok ahead of upcoming races. Platforms that promised faster takedowns of synthetic video and fabricated quotes are finding the volume keeps outpacing their review teams, and regulators who might otherwise step in are dealing with their own technical headaches.
The newsletter's lead item today sits elsewhere, on a stranger wrinkle in the AI economy: failed startups selling off their data troves to training labs hungry for anything that looks like real-world material. Keelan Miskell, who wound down a crypto startup, says he's made tens of thousands of dollars offloading its codebase and internal records this way. That trade, and a separate piece on researchers chasing animal communication through AI, round out an issue that keeps circling back to the same tension: systems built to move fast are colliding with oversight that can't keep pace, whether that's election watchdogs or the labs buying up digital leftovers.
When Keelan Miskell began winding down his crypto startup, he didn’t think its data was worth much. But he’s since turned the company’s digital archives into tens of thousands of dollars, selling its codebase and workspace data.
Why this matters The Guardian's reporting on platforms failing to contain AI-generated election lies lands alongside a separate story about hackers knocking out the FEC's website during what Gizmodo calls the costliest midterms on record. Axios adds software flaws in voting machines to the pile. None of this is coincidental timing so much as a sign that the infrastructure around elections, technical and informational, is buckling at the same moment generative tools make fabricated content cheaper to produce and harder to trace.
For builders in this space, the uncomfortable detail is the Keelan Miskell item: AI labs are now buying up data from dead startups, crypto ventures included, to train models. That's a pipeline almost nobody is auditing. If training data can come from a defunct crypto company's digital archives, the provenance question for any model touching political content gets murkier, not clearer.
The "blueprint for AI strengthening democracy" piece reads less like a solution and more like a wish list until someone addresses where the training data and the platform moderation failures actually meet.
Related Reading
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
From Archives Akamai adds AI Inference and Agent Sandbox to complete its full AI stack
From Archives Bengaluru Startup Builds AI Agents for Existing Hardware Design Tools
Common Questions Answered
Why are social media platforms struggling to contain AI-generated election lies?
Social media platforms like Facebook, X, and TikTok are finding that the volume of AI-generated falsehoods spreading across their networks is outpacing their review capabilities, despite promises for faster takedowns of synthetic video and fabricated quotes. The platforms are unable to keep up with the speed and scale at which generative AI tools are making fabricated content cheaper and easier to produce during election cycles.
What recent incident affected the Federal Election Commission's ability to monitor election interference?
The Federal Election Commission's website went dark for parts of the week after hackers disrupted it, according to MIT Technology Review's The Download newsletter. This outage occurred at a particularly problematic time when the agency was already stretched thin trying to track AI-generated falsehoods spreading across social media platforms ahead of upcoming races.
How are dead startups contributing to AI development according to the article?
Entrepreneurs like Keelan Miskell are monetizing their defunct crypto startups by selling their digital archives, including codebase and workspace data, for tens of thousands of dollars. This startup data is being repurposed to fuel AI development and other technological applications, demonstrating an unexpected revenue stream from failed companies.
What multiple infrastructure vulnerabilities are threatening election integrity?
The article highlights that election infrastructure is buckling under pressure from multiple simultaneous threats, including AI-generated election misinformation spreading on social media, hackers disrupting the FEC's website, and software flaws discovered in voting machines. These technical and informational vulnerabilities are converging during what Gizmodo reports as the costliest midterms on record.
Further Reading
- A.I. and Its Dangers Are Embedded In the Midterm Elections - The New York Times
- AI disinformation is outpacing platform moderation before the US midterms - The Next Web
- Experts warn of AI disinformation ahead of US elections - The Guardian
- The Second AI Election: Testing the Safeguards Before the 2026 Midterms - The Fulcrum
- Bots and Ballot Boxes: What Today's Technology Means for Election Integrity - The Leadership Conference on Civil and Human Rights
