There aren’t protocols and regulations in place for how to manage algorithms and data in many Western countries, but they’re coming—and social media companies hoping to push that day off by opening their books and auditing one another are only slowing what’s to come.
Midterm elections could well determine the future of algorithmic regulation, Justin Sherman, a geopolitics and internet expert and nonresident fellow with the Atlantic Council, told IT Brew. With a new Congress on the horizon, who controls the House and Senate will likely be the main factor in what the laws look like.
“At this point, it’s considered inevitable that there will be some kind of algorithmic regulation,” Sherman said. “Whether that’s something as simple as saying, ‘These companies need to give more data to researchers in a more open way to study the platforms,’ or whether that’s imposing things like mandated third-party independent algorithmic audits, which is a growing area of focus.”
Regulators, mount up. The US has long had a contentious relationship with social media companies. Both sides of the aisle have found reasons to go after Facebook, Twitter, and Google.
In a move responsive to concerns over the platform’s power over elections and political news, Facebook announced in August that it would restrict political advertising in the US from Nov. 1 to Election Day; a prohibition that includes “ads that call into question the legitimacy of the methods and processes of elections, as well as their outcomes.”
TikTok has been the subject of perennial accusations that it may present a national security threat. The Biden administration and the company are currently in talks to work out a deal for TikTok to remain working in the US, but any challenges at the federal level are likely to run up against constitutional issues.
To sweeten the deal for Biden—and to try to stay above the fray—TikTok is denying national politicians a platform on the service during the midterms, following Facebook’s lead. Both companies are gambling on staying on good terms with Washington, no matter who’s in charge, in an effort to forestall regulations on their algorithms and business models.
Top insights for IT pros
From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.
But that likely won’t be enough to assuage concerns over what the companies are doing to society. The algorithms tend to result in a push toward extremism, explained Sherman. That’s clearly a problem, and it’s one that puts the US government in a sticky position: balancing dealing with a toxic social issue with seeming to side with one end of the political aisle.
“As these discussions move towards the issues of black boxes and understanding algorithms, that same concern keeps popping up,” Sherman told IT Brew.
Under the hood. Machine learning has increasingly taken a major role in how social media companies manipulate user experience. As Max Fisher, author of The Chaos Machine, a study of social media, recently told IT Brew, what the algorithm does is more important than how. He used YouTube as an example.
“These machine learning systems…can train the system to increase watch time or increase certain types of views,” Fisher said. “But we don’t know exactly how the system goes about maximizing that and what sort of content it might choose to promote.”
Black box systems are incredibly complex, Jay Wang, a research student at the Georgia Institute of Technology who specializes in the study of black box algorithm technology, told IT Brew.
“Anyone in Facebook doesn’t know how Facebook works, it’s some kind of magic machinery model—they feed in a huge amount of data and the model does some magic, and they can produce a recommendation for physical media,” Wang said. “But no one knows how the algorithm works.”—EH
Do you work in IT or have information about your IT department you want to share? Email [email protected] or DM @EoinHiggins_ on Twitter.