It can be hard to remember from this side of a successful foreign intervention in a US presidential election, but in April 2016, Mark Zuckerberg was widely praised when he spoke out – however obliquely – against Donald Trump’s then longshot presidential candidacy.

“I’m starting to see people and nations turning inward against this idea of a connected world,” the Facebook chief executive said in a speech. “I hear fearful voices talking about building walls … It takes courage to choose hope over fear.”

Three and a half years later, another US presidential campaign is well under way and courage is not exactly thick on the ground at Facebook’s headquarters in Menlo Park. For the past few weeks, the company has been embroiled in a slow-motion scandal over its inexplicable policies on political misinformation, one that epitomizes the kind of moral cowardice that characterizes Facebook’s approach to its political challenges.

The controversy kicked off last month when the Facebook executive Nick Clegg revealed in a speech that politicians are exempt from the company’s third-party fact-checking program. The policy drew immediate scrutiny due to the distinctly dystopian context in which the speech was made: just as the president of the United States was actively promoting misinformation on Facebook to smear a political rival and undermine an impeachment inquiry.

Trump’s Facebook page has launched 5,883 different ads since news of the Ukraine-call whistleblower broke on 18 September, 40% of which mention impeachment, according to a Guardian analysis of the Facebook political ad archive. The Trump campaign has spent between $1.3m and $3.8m promoting the impeachment ads, which had been viewed between 26.5m and 43.9m times, as of Wednesday. (The ad archive reports spending and impression data as a range.)

Most of the ads employ the Trump campaign’s usual brand of rhetoric: railing against the Democratic “witch-hunt” and “fake news” media; singling out non-white politicians for attacks (the Texas congressman Al Green is referred to in one set as “Low-IQ”); and encouraging a sense of grievance while seeking donations (“They want to take YOUR VOTE away”).

But one of the ads has caused particular concern because it is blatantly misleading. The 30-second video falsely claims that Joe Biden “promised Ukraine a billion dollars if they fired the prosecutor investigating his son’s company”. It uses audio of Biden speaking loosely of his role in Barack Obama’s foreign policy – “If the prosecutor is not fired, you’re not getting the money,” he says, “Well, son of a bitch, he got fired” – to create the impression that Biden admitted to extorting Ukraine on behalf of his son. CNN declined to run the ad, saying: “In addition to disparaging CNN and its journalists, the ad makes assertions that have been proven demonstrably false by various news outlets, including CNN.”

Facebook’s insistence on protecting Trump’s right to mislead the public and distort verifiable facts is a strange look for a company that attempted to burnish its reputation post-Russian interference with an ad campaign declaring “Fake news is not your friend”. And attempts by the company to justify the decision are nearly impossible to parse.

In a letter to the Biden campaign, which complained about the Trump ad, Facebook’s elections policy director, Katie Harbath, invoked “Facebook’s fundamental belief in free expression, respect for the democratic process, and belief that, in mature democracies with a free press, political speech is already arguably the most scrutinized speech there is”. She then explained that if a politician shared a link to a viral hoax written by someone else, that post would be rejected as an advertisement, but if the same politician wrote the viral hoax in his own words, that would not only be fine, but Facebook would not allow third-party fact checkers to flag the information as false.

When I asked Facebook whether this new rule would also exempt politicians who promote vaccine misinformation, which is explicitly barred from advertisements, a company spokeswoman said that it would not. If your head is spinning, you are not alone. These rules do not make any sense.

What’s even more absurd is that Facebook actually did take action against the ads, just for a different reason. As I reported on 2 October, the Biden ads were briefly blocked for violating Facebook’s policy against profanity. (After the Trump campaign bleeped “son of a bitch”, they were allowed to run again.)

Why does Facebook’s fundamental belief in free expression privilege the right of a politician to lie but not to use a swear word?

It has been tempting for some to ascribe Facebook’s motives here to economic self-interest. The Trump campaign is a major advertiser on Facebook, after all, much larger than the Biden campaign.

But I suspect that this incident has less to do with revenues and more to do with Zuckerberg’s fundamental political programming, which defaults to protecting incumbent power above all else.

Was it really an act of “courage” for Zuckerberg to criticize Trump’s xenophobia in April 2016, when a Hillary Clinton coronation appeared assured?

Over and over again, Zuckerberg and Facebook have failed to take action to prevent real and lasting harm, when that action involved challenging the status quo or state power. This was demonstrated most devastatingly by its unwillingness to rein in Myanmar’s military until after a brutal campaign of ethnic cleansing against the Rohingya had already been accomplished, as well as in its foot-dragging on banning white nationalism. This predilection for the powerful is how Zuckerberg went from calling out Trump on building walls to an Oval Office grip and grin.

We are living in an age when everyone from cheap furniture makers to video game publishers is being forced to pick sides in serious struggles between business and barbarism, between authoritarianism and democracy. That Zuckerberg refuses to do so reveals whose side he’s really on.

• Welcome to Dispatches from Silicon Valley, a new weekly column from me, Guardian US technology reporter Julia Carrie Wong. Each week, I’ll be bringing you reporting and analysis from the epicenter of the tech industry. If you have ideas or comments for future dispatches, email me at julia.wong@theguardian.com.