Tech Platforms Shape Voter Choices as AI Risks Grow

Artificial intelligence now plays a role across every major social media platform for the 2026 midterm elections. Chatbots, media-generation services and automated content moderation increasingly shape voter choices. Technology companies have introduced tools that can put obstacles between voters and reliable election information, while addressing the risks inconsistently or not at all.

Who Shapes Election Information

During the 2024 national election, platforms offered basic AI image-generation tools. By 2026, those capabilities had grown into advanced services that combine audio, video and image functions. Political experts and new research cited in The New York Times report say some tools arrived only months before the election.

Researchers said AI tools have contributed to biased, inaccurate information about polls. They’ve also amplified foreign influence operations. The findings identify platforms and their AI services as channels bringing election information to voters, but don’t say how many people used the tools or encountered inaccurate material.

Researchers have also found that the tools elevate low-quality content, which they call “slop.” That material taints the training data chatbot responses draw on. The account connects the spread of low-quality content to what chatbots use, but doesn’t measure its effect on specific responses or voters.

What Companies Have Disclosed

Lawmakers and watchdog groups have pressed technology companies for details about testing, safeguards and voter interactions with AI. They say companies haven’t shared enough. Their requests ask how companies tested the tools, what protections they have in place and how extensively voters use AI services during the election.

“We’re in uncharted territory at this point,” said Deen Freelon, a professor of communication at the University of Pennsylvania who studies digital politics. His comment captures uncertainty about AI’s role in the election, and the report says some services reached platforms only months before the vote.

The account names requests for information and company responses, but identifies no specific law or enforcement action aimed at the platforms. Nor does it provide company-by-company descriptions of safeguards or testing results. Limited disclosures leave lawmakers and watchdog groups without the information they’ve sought about the systems’ preparation and reach.

The Limits of the Record

The report describes technology companies rapidly integrating AI across major social media platforms. It identifies potential obstacles to reliable election information, inaccurate poll material, foreign influence operations and low-quality chatbot training data. It gives no figures for company revenue or profits tied to these tools. It also says nothing about job losses, worker conditions, union activity or organized labor’s response, and names no strike or labor campaign related to the platforms.

The report describes a political response: lawmakers and watchdog groups pressing companies for details about testing, safeguards and voter interactions. Those requests haven’t produced the level of disclosure the groups say they need. The report gives no measure of how many voters encountered AI-generated election content or how the platforms’ services changed individual voting decisions.