April 10, 2018 5 min read

At 2:15 p.m. ET Tuesday, Facebook CEO Mark Zuckerberg is scheduled to take the stand before the joint Senate Judiciary and Commerce Committees. The following day, he’ll testify before the House of Representatives Energy and Commerce Committee. He plans to wear a suit and take accountability for Facebook’s “big mistake,” and his seven-page testimony has been released by the House. Zuckerberg’s testimony will be streamed live on PBS NewsHour here, and if you’d rather watch it on Facebook itself, C-SPAN 3 will have another live stream here.

Ahead of his hearing, we pulled apart Zuckerberg’s testimony document. First, here are recaps of the two significant events that landed Facebook in hot water.

Recap: How did Cambridge Analytica get access to so much user data?

About 87 million is an oft-cited figure in connection with the number of users whose data was mishandled by Cambridge Analytica, a political data firm hired by President Donald Trump’s campaign before the 2016 election. But the story behind how the firm got its hands on that information is a little murkier.

Zuckerberg’s testimony claims that Cambridge University researcher Aleksandr Kogan created a personality quiz app on the platform in 2013 and that about 300,000 people installed it. At the time, laxer privacy settings across Facebook meant Kogan had access to data from tens of millions more users after their friends had installed the app. Zuckerberg says Facebook wasn’t aware that Kogan had shared data with Cambridge Analytica until 2015 and that the company “ultimately” formally certified that they had deleted “all improperly acquired data.”

However, the firm likely used the information to build psychological profiles to categorize individual American voters. Besides a claim, there’s no clear evidence Cambridge Analytica has since deleted that data — but Zuckerberg says they’ve agreed to a forensic audit by a firm Facebook hired.

Recap: How did Russian agencies spread misinformation ahead of the 2016 U.S. election?

After the election, Facebook found that “bad actors” ran ads and coordinated groups of fake accounts to endorse or denounce certain causes and candidates — or sow widespread distrust in the political institutions themselves. One “disinformation campaign” linked to about 80,000 Facebook posts in two years was run by the Russia-based Internet Research Agency (IRA), and the company estimates that 146 million people might have been served its content on Facebook and Instagram combined. The IRA also invested about $100,000 in Facebook ads seen by about 11 million people in the U.S., according to Zuckerberg’s testimony.

Here are six concrete takeaways from Facebook’s plan to prevent future incidents.

1. More secondhand data protections

The company is introducing restrictions on tools such as “groups” and “events” to protect the data of other users besides the one signing in, according to Zuckerberg’s testimony. There’s no word yet from Facebook on exactly how those tools will be restricted, and the company didn’t respond to a request for comment by publication time.

2. Restricted data permissions for apps

For Facebook users who haven’t used an app in three months or more, Zuckerberg says the app developers will no longer have access to their data. And instead of granting apps access to data such as public profiles and friends lists, users will now only need to provide their names, profile photos and email addresses. Facebook claims it has already begun showing all users a list of the apps they’ve used and clear ways to revoke their data permissions at the top of the News Feed. As of publication time, Entrepreneur staffers were unable to see this tool.

3. More oversight for app developers

Zuckerberg says there will be more regulation on the back end as well, with developers signing contracts with “strict requirements in order to ask anyone for access to their posts or other private data.” There’s no word yet from Facebook on what those requirements will be.

4. More oversight for advertisers

One key change coming to Facebook ads: Every advertiser behind issue or political ads will need to be authorized based on identity and location. The company plans to extend that verification to users who manage large pages, in the hopes of preventing viral misinformation.

5. More political ad transparency

In the coming months, labels will be introduced for political and issue ads, including who paid for them, and Facebook will launch a “searchable archive” of past political ads. The company is also testing a tool that shows all of the ads any given page is running and plans to launch it worldwide in summer 2018.

6. Quicker implementation timeline

Zuckerberg says all Facebook ad changes should be complete months before the 2018 elections, as well as for upcoming elections over the next year in India, Pakistan, Mexico and Brazil.