So how can this sort of meddling be prevented in the future? Under its terms of service, Facebook already has the legal authority to delete any account created with “false personal information.” Hence, Facebook has the means to kick off those who use its platform under an alias, hiding the fact that they represent a foreign entity. In its announcement about the Russia-linked political ads, Facebook contended it would deploy technological solutions to “help keep activity on Facebook authentic.”

I have no doubt that Facebook’s data scientists are hard at work on this solution. But this represents an incomplete response at best. What Facebook didn’t suggest, and what it and other social media companies routinely fight tooth and nail, is a regulatory response. Let “machine learning” solve the problem, says Facebook, not new laws and opportunities for review by outsiders.

Social media audiences benefit from knowing the source of an ad or post. Presumably, Facebook users would be better able to judge the veracity of a political message if they knew it came from George Soros, the Koch brothers, or a Putin-friendly troll farm in Russia. If we want to avoid similar problems in 2020, the law needs to update political advertising disclosure rules for the digital age.

First, there needs to be reform of the Federal Election Commission’s so-called “Internet exemption.” In 2006 rulemaking, the commission choose to immunize content that is posted online for free from regulation. The intent then was to protect bloggers, who were viewed as individual citizens sharing their voices on political issues that shouldn’t have to report their activities to the government. But now we live in an age of social media influencers who routinely post content on social media platforms like Facebook and Twitter, often for big paydays.

It won’t be easy, but the Internet exemption should be rewritten to address digital political advertising. Sometimes sophisticated ads are released on social media through back channels. The hope is that these ads will go viral, being transmitted around the web for free as people become either enthused or outraged by them. Some steps need to be taken to make sure that this kind political astroturfing isn’t allowed to escape the disclosure rules routinely applied to paid ads.

Second, Facebook needs to disclose the bulk of the data about who it sells political advertising space to and who these ads reach. At this point, Facebook has taken a hard line, refusing to turn over this information and contending that political advertisers and commercial advertisers deserve the same amount of total privacy. Even though the stakes are surely higher with political ads—potentially influencing the outcome of a presidential election—Facebook refuses to see a difference between selling candidates and selling soap.

Yet there are good reasons for requiring Facebook to treat political advertising differently. Armed with information about the frequency of political ad buys, their cost, their audiences, and their content, researchers could know more about the information that shapes modern political discourse. It’s easy enough to catalog television advertisements in this regard, but political advertising on Facebook remains a black box. Political ads on Facebook can evaporate once they are screened on a user smartphone, leading critics to describe them as “dark ads.”

All this goes to the larger question of what kind of entity Facebook is and how we should treat it. Just like television networks and newspapers, Facebook makes choices about what we see. Yet unlike television stations or newspapers—where we can see who the decision makers are and what decisions they make—Facebook keeps things hidden. It won’t reveal how it decides what enters our news feeds. It won’t even reveal the content of these bogus political ads from Russian backers. It is time for lawmakers to force Facebook to be more transparent.

Mark Bartholomew is a professor at the University at Buffalo School of Law and the author of Adcreep: The Case Against Modern Marketing.