A recent study by Global Witness, a non-governmental organization, has uncovered substantial evidence of far-right political bias in the algorithmic recommendations of social media giants TikTok and X in Germany, just days before the country's federal election. The research, which analyzed the content displayed to new users via algorithmically sorted 'For You' feeds, found that both platforms skewed heavily towards amplifying content that favors the AfD party.

On TikTok, the study revealed that a staggering 78% of the political content algorithmically recommended to its test accounts, which did not follow any specific political parties, was supportive of the AfD party. This figure far exceeds the level of support the party is achieving in current polling, where it attracts backing from around 20% of German voters. X, formerly known as Twitter, displayed a similar bias, with 64% of recommended political content supporting the AfD.

The research also tested for general left- or right-leaning political bias in the platforms' algorithmic recommendations. The findings suggest that non-partisan social media users in Germany are being exposed to right-leaning content more than twice as much as left-leaning content in the lead-up to the country's federal elections. TikTok displayed the greatest right-wing skew, showing right-leaning content 74% of the time, while X was not far behind, with 72%.

Meta's Instagram was also tested and found to lean right, although the level of political bias it displayed was lower, with 59% of political content being right-wing. The researchers set up test accounts on each platform, following the accounts of the four biggest political parties in Germany, along with their respective leaders' accounts. They then manually collected and analyzed the content each platform pushed at the test accounts, finding a substantial right-wing skew in what was being algorithmically pushed to users.

Ellen Judson, a senior campaigner at Global Witness, expressed concerns over the lack of transparency from platforms about how their recommender systems work. "We found this evidence that suggests bias, but there's still a lack of transparency from platforms about how their recommender systems work," she said. Judson attributed the bias to the unintended side effect of algorithms designed to maximize user engagement, which can conflict with public interest and democratic objectives.

The findings are consistent with other social media research conducted by Global Witness around recent elections in the U.S., Ireland, and Romania. Similar studies have also found evidence that social media algorithms lean right. However, social media firms typically try to downplay allegations of algorithmic bias. TikTok suggested that the researchers' methodology was flawed, arguing that it was not possible to draw conclusions of algorithmic bias from a handful of tests. X did not respond to the findings.

The European Union's Digital Services Act (DSA) aims to improve transparency and accountability around algorithmic decision-making on major platforms. However, the regulation has yet to deliver quick results in addressing concerns around social media and democratic risks. The EU's approach may also ultimately be shown to be too cautious to move the needle as fast as it needs to move to keep up with algorithmically amplified threats.

Global Witness has shared its findings with EU officials responsible for enforcing the bloc's algorithmic accountability rules on large platforms. The organization is hoping that the Commission will investigate whether there is political bias and address it if necessary. The pan-EU regulation empowers enforcers to levy penalties of up to 6% of global annual turnover for infringements, and even temporarily block access to violating platforms if they refuse to comply.

The study's findings raise important questions about the role of social media in shaping public opinion and influencing democratic outcomes. As the EU continues to grapple with the challenges of regulating social media, it remains to be seen whether the DSA will be effective in promoting transparency and accountability around algorithmic decision-making.