Buck targets social media companies

Bill aims to force businesses to be more transparent with algorithms it’s a good business more recent

Social media and other internet platforms rely on algorithms to curate what users see based on information the companies collect as people scroll through their feeds.

U.S. Rep. Ken Buck has introduced in a bill in the U.S. House aimed at curbing that.

Buck, who represents Colorado’s 4th Congressional District, is sponsoring a bipartisan companion bill to one in the U.S. Senate, dubbed “The Filter Bubble Transparency Act,” that would require companies to alert users the companies use algorithms that rely on personal data. The bill would require companies to give users the option to see content without those algorithms, and to switch back and forth. A similar bill was introduced in 2019.

“I guess practice if you’re Facebook or Google, but what they’re trying to do is they’re trying to increase engagement on their platforms so that they can advertise to people. The challenge is that we end up with a more polarized citizenry and really a country that knows less about what other people think on an issue and therefore have …. less understanding, less patience maybe for other thoughts because all they’ve been fed are their own views over and over again,” Buck said in an interview Tuesday.

Some social media platforms such as Facebook allow users to change their feeds and view them in chronological order, rather than just what’s targeted at them, but users have to select this option every time they want to see their recent posts.

Buck wants consumers to have the choice to view content not just chronologically at all times, but without what he refers to as manipulation.

“With Google, if you put in search results, and you don’t have an algorithm, you’re going to get back one set of results. If you do have an algorithm based on what you’ve opened in the past, you’re going to get back another set,” he said.

Casey Fiesler, who teaches and researches technologies and online communities as a fellow for the Silicon Flatirons Institute for Law, Technology and Entrepreneurship, said she worries about highly personalized feeds contributing to confirmation bias. For example, she often heard in the wake of the last election people saying, “I haven’t heard of a single person who voted for Joe Biden,” and alternately, “I don’t know a single person who voted for Donald Trump.”

Thirty years ago, if someone said that about a specific candidate during an election, they might have also understood that the number of people they are connected with is not very large or representative of the population as a whole.

“But when you’re on social media, it feels like your world is bigger. It feels like you’re seeing more perspective because you’re scrolling through thousands and thousands of posts on Facebook, including stuff from people you don’t follow, if you follow groups or it’s recommending content … ” Fiesler said.

Still, the professor wonders how long users would like not having that curation, which shows them content they would be interested in seeing, particularly on platforms like Tiktok.

Buck’s bill provides exemptions: companies would still be able to restrict content for children and it would not apply to companies that collect data from fewer than 1 million people, did not have more than 500 employees in the most recent six-month period and averaged less than $50 million annually in gross receipts over the three-year period.

It also wouldn’t apply to companies doing research that isn’t for profit or other types of algorithms.

The Chamber of Progress, a tech industry group that calls itself center-left, released a statement opposing the legislation, saying minsinformation, spam and hate speech will be allowed to spread online if algorithms are removed.

It cited a Morning Consult survey, showing a majority of users wanting stricter content standards on Facebook.

But Buck dismissed those complaints. He said the bill would not affect protections already in place for all users and only applies to algorithms targeting some users based on specific interests or personal data collected about them.