Trump Exploited Facebook. Now Facebook Protects Him.

In early 2025, following Trump’s “Liberation Day” tariff announcements, I tried to run ads for a client that provides daily market analysis to financial advisors and investors. The content was straightforward: analysis of how tariffs might affect the economy, with a call to subscribe for ongoing updates. The kind of commentary Bloomberg or CNBC runs constantly.

Google disapproved the ads. The reason: “Election advertising in the United States. Certificate required.”

I appealed. The appeal failed. I appealed again. Failed again. Between May 19 and June 9, 2025, I filed five consecutive appeals, each reviewed by a human, each denied. The ads mentioned Trump’s tariffs. Under current platform policy, that constitutes election advertising.

On Meta, the same content triggered the same flags. To run anything at all, I had to categorize my client’s market commentary as “Social issues, elections or politics,” accepting reduced targeting options and additional restrictions. We were not weighing in on any election. There was no election. Trump had won just a few months earlier, and the next presidential race was more than three years away. But under the rules these platforms built after 2016, mentioning a sitting president in any commercial context requires election advertising certification, regardless of whether the content has anything to do with voting.

This would not have happened during the Obama years. It would not have happened during Biden’s term in the same way. The regulations that created this situation emerged from the 2016 election fallout, and the timeline of their adoption has produced an asymmetry that benefits the current administration in ways the platforms may not have intended, but have done nothing to correct.

On March 17, 2018, The Guardian and The New York Times published their Cambridge Analytica exposés.¹ Christopher Wylie revealed that Cambridge Analytica had harvested data from up to 87 million Facebook profiles for political microtargeting, including for the Trump 2016 campaign.² Within weeks, both Google and Meta announced new verification requirements for election advertising.

Google’s initial policy, announced May 4, 2018, required verification for ads featuring “federal candidates or current elected federal officeholders.”³ Enforcement began July 10. Meta launched its authorization process for political advertisers that same month, requiring identity verification and “paid for by” disclosures on all political and issue ads.⁴ Both platforms framed these changes as responses to foreign interference and data misuse, efforts to restore trust in the democratic process.

The threats were real. Russian operatives had purchased ads on Facebook to sow division during the 2016 election.⁵ The Internet Research Agency had created fake American personas to spread disinformation at scale.⁶ Cambridge Analytica had exploited platform vulnerabilities to harvest data for political targeting without user consent. Some regulatory response was warranted, and the platforms faced legitimate pressure to prevent their advertising infrastructure from being weaponized against democratic processes.

The beneficiaries of that exploitation are a matter of public record. Cambridge Analytica used the harvested data to build psychographic profiles for the 2016 presidential campaigns of Ted Cruz and Donald Trump. Federal Election Commission filings show Cruz’s campaign paid the firm $5.8 million between July 2015 and June 2016; Trump’s campaign and affiliated political action committees paid approximately $5.9 million.⁷ Facebook confirmed in April 2018 that 87 million users had their data harvested, with 70.6 million of those users located in the United States.⁸ The same data operation that triggered the platform restrictions now shielding Trump from commercial criticism was the data operation his campaign paid millions to deploy.

But the definitions kept expanding beyond foreign interference. In November 2019, Google extended the policy to cover state-level officeholders, with enforcement beginning December 3.⁹ Meta’s definition grew to encompass not just candidates and officeholders but “social issues” broadly, including “the economy, or civil and social rights.”¹⁰ The platforms built systems to catch foreign actors manipulating elections and ended up catching anyone who mentioned the president while trying to sell something.

Barack Obama governed for eight years with essentially no such restrictions on commercial speech about his administration. Advertisers could run content mentioning Obama without triggering special verification requirements, without having their appeals denied, without being forced to categorize market commentary as political content. The same was largely true for Biden, who inherited these restrictions when he took office in 2021 but did not govern as a perpetual campaigner. The distinction between his administration and his electoral prospects remained relatively clear, even if the platform policies did not recognize it.

Trump’s second term operates in a completely different environment. Any ad mentioning him triggers election advertising requirements, full stop. Google’s current policy explicitly states that “US Election Ads are ads that feature any of the following: A current officeholder or candidate for an elected federal office. This includes federal offices such as that of the President or Vice President.”¹¹ The policy makes no distinction between campaign advocacy and economic commentary. Mention the president, and you are running an election ad.

Automated review systems, powered by machine learning, now scan ad content for keywords and flag anything that mentions a sitting president, a political party, or a policy associated with either. The technology built to catch foreign interference at scale now catches domestic financial commentary with the same broad net. A video mentioning “Trump tariffs” triggers the same automated response as a video urging people to vote for Trump.

Human review exists, but it operates within the same framework. When I appealed my disapprovals, actual people reviewed the appeals and upheld them. They were not evaluating whether my content was actually electoral in nature. They were evaluating whether it fell within the policy definition, which says that any ad featuring a “current officeholder” is an election ad. By that standard, my appeals were correctly denied. The policy itself is the problem.

The obvious counterargument is that advertisers can simply get certified. Submit identification, complete the verification process, run the ads with appropriate disclosures. But compliance creates its own problems. Platforms require documentation linking your legal entity name to any trade name you advertise under. Many businesses, particularly those that have built authority under a specific brand, operate with “doing business as” structures that lack the precise documentation platforms demand. Your ads must then display your legal entity name rather than the brand your audience recognizes. For a financial analysis service that has spent years building trust under a particular name, having ads suddenly attributed to an unfamiliar holding company erodes the authority that makes the content valuable. The path to compliance exists, but for many legitimate advertisers, it leads to a situation they cannot practically execute.

This acceleration in automated enforcement has made the restrictions both broader and less visible. There is no public announcement when the algorithm gets more aggressive. There is no notice when the keyword list expands. Advertisers simply discover, through rejection notices, that content they ran successfully last month no longer clears review this month.

The platforms did not design these policies to benefit one administration over another. They were responding to genuine threats, and the pressure to act was intense. But the timeline of implementation produced an asymmetric outcome regardless of intent. A financial analyst who wanted to run ads about Obamacare’s economic impact in 2012 faced no verification requirements. A financial analyst who wants to run ads about Trump’s economic impact in 2025 must either register as an election advertiser, accept severe targeting restrictions, display disclosures that undermine their brand, or abandon the effort entirely.

The campaigns that most effectively exploited the previous system now receive the greatest protection under the new one.

Trump has also collapsed the distinction between governing and campaigning more thoroughly than any modern president. His policies are framed as extensions of his political movement. His tariffs are “Liberation Day.” His economic decisions are presented as fulfillment of campaign promises. The platforms built rules that assume mentioning a sitting president is inherently electoral, and Trump has made that assumption functionally true for himself. The regulations reflect his behavior but protect him from criticism.

In January 2025, days before Trump’s second inauguration, Mark Zuckerberg announced that Meta would end its third-party fact-checking program in the United States, citing the political environment and a desire to return to “free expression.”¹² The same platforms that require election advertising verification for financial market commentary decided that fact-checking political claims was too restrictive.

The apparent contradiction reveals the platforms’ actual priorities. They are not consistently pro-speech or anti-speech. They are pro-liability-management. The platforms implemented election advertising restrictions after 2016 because they faced intense scrutiny over their role in foreign interference and data misuse. The fact-checking rollback came in 2025 because the platforms faced intense pressure from an incoming administration that had made its hostility to content moderation clear.

The result is a system that creates friction for ordinary advertisers while removing friction for political actors. A financial analyst needs election certification to mention tariffs. A political operative needs only to wait for the fact-checkers to disappear.

Those of us who work in advertising see these restrictions from the inside. We see the rejection notices, the failed appeals, the forced categorizations. We know that content about economic policy now requires the same bureaucratic overhead as campaign ads. We know that previous presidents did not benefit from this kind of insulation. We also know these platforms were responding to real threats when they built these systems, which makes the asymmetric outcome harder to dismiss and harder to fix.

The platforms will not correct this on their own. They have no incentive to. Whether the rest of us are paying attention is another matter.

Trump is the only modern president to serve non-consecutive terms, and that timing matters. His first campaign paid millions for data harvested from 87 million users without consent. The platforms built restrictions in response. His second administration now benefits from those restrictions while media companies settle lawsuits they could win, platforms abandon fact-checking, and corporations gut internal policies that might produce critical coverage. He exploited the old system, triggered the new one, and now operates within a media environment that has capitulated at every level. No other president has occupied this position because no other president served across both eras.

What can you do about it?

Go local. Soft Secession is the framework where you can push state-level governance to regulate social media platforms. It’s already been done by California and we can go further. To learn more about what it takes to hold these corporations accountable and actually win against fascism, check out TheExistentialistRepublic.com for the Intro to Soft Secession primer booklet and BuyMeACoffee.com/TheER to get free activist training materials.

References

1. Cadwalladr, C., & Graham-Harrison, E. (2018, March 17). Revealed: 50 million Facebook profiles harvested for Cambridge Analytica in major data breach. The Guardian.

2. Rosenberg, M., Confessore, N., & Cadwalladr, C. (2018, March 17). How Trump consultants exploited the Facebook data of millions. The New York Times.

3. Walker, K. (2018, May 4). Supporting election integrity through greater advertising transparency. Google: The Keyword.

4. Goldman, R. (2018, May 24). Shining a light on ads with political content. Meta Newsroom.

5. United States Senate Select Committee on Intelligence. (2019). Russian active measures campaigns and interference in the 2016 U.S. election, Volume 2: Russia’s use of social media.

6. United States Department of Justice. (2018, February 16). Grand jury indicts thirteen Russian individuals and three Russian companies for scheme to interfere in the United States political system.

7. OpenSecrets. (2018, March 23). Before Trump, Cambridge Analytica was on team Cruz.

8. Kang, C., & Frenkel, S. (2018, April 4). Facebook says Cambridge Analytica harvested data of up to 87 million users. The New York Times.

9. Google. (2019, November). Update to policy on election advertising in the United States. Google Ads Help.

10. Meta. (n.d.). Ads about social issues, elections or politics. Meta Transparency Center.

11. Google. (n.d.). Political content: United States (US) Election Ads. Google Ads Help.

12. Zuckerberg, M. (2025, January 7). More speech and fewer mistakes. Meta Newsroom.