Buying Elections: Meta's Political Ad System and the Death of Democratic Discourse

In every major democracy on earth, Meta's advertising platform has become the primary battlefield for political campaigns. Political ad spending on Meta platforms exceeded $4.5 billion during the 2024 US election cycle alone, dwarfing spending on television, radio, and all other digital platforms combined. This dominance would be concerning even with robust transparency and oversight. In reality, Meta's political advertising system offers campaigns unprecedented ability to micro-target voters with tailored messages while providing regulators and the public with only a partial view of what's being said, to whom, and by whom.

The Micro-Targeting Problem

The core issue with political advertising on Meta isn't volume — it's precision. Meta's ad targeting allows campaigns to segment voters into thousands of distinct audiences based on demographics, interests, behaviors, and custom data lists uploaded by the campaigns themselves. A political candidate can show a pro-labor message to union members in Michigan, a pro-business message to executives in Connecticut, and an anti-immigration message to rural voters in Texas — all simultaneously, with no audience ever seeing the other messages. This capability fundamentally undermines democratic discourse, which depends on candidates making public commitments that all voters can evaluate. When every voter sees a customized message, the concept of a public campaign platform becomes meaningless.

Transparency Theater

How does your site score?

Run a free scan and get actionable improvement prompts in 30 seconds.

Meta launched its Ad Library in 2018 in response to the Cambridge Analytica scandal, promising transparency for political advertising. But the Ad Library has significant gaps. Academic researchers at NYU's Cybersecurity for Democracy project found that the library captures only 60-70% of ads that qualify as political, missing issue ads that discuss policy topics without explicitly naming candidates. More critically, the Ad Library reveals nothing about targeting criteria — the public can see that an ad existed but cannot see the demographic, geographic, and behavioral parameters that determined who saw it. Without targeting data, the Ad Library provides a catalog of messages without the context needed to understand their strategic purpose or identify manipulation.

Foreign Interference Continues

Despite Meta's post-2016 promises to prevent foreign interference in elections, the problem persists. Researchers at the Stanford Internet Observatory identified over 2,000 Facebook and Instagram accounts engaged in suspected foreign political influence operations in 2025, including networks linked to Russia, China, and Iran. Meta's identity verification requirements for political advertisers have proven easy to circumvent using shell companies, local intermediaries, and stolen identity documents. The company removes these networks when researchers identify them — often months after the ads have already run and reached millions of voters — but the reactive approach means foreign actors consistently outpace Meta's detection capabilities.

The fundamental reform needed is straightforward: require full targeting transparency for all political advertising, mandate real-time disclosure of ad spending above a threshold amount, and impose meaningful penalties for platforms that fail to prevent foreign political ad purchases. Multiple legislative proposals have incorporated these requirements, but Meta's lobbying operation — which spent over $20 million in 2025 — has been effective at preventing their passage. Until these reforms are enacted, Meta will continue to be the most powerful and least accountable tool for shaping democratic elections in the modern world.