Ofcom: UK Media Plurality At Risk From Online Algorithms

Ofcom have found that online algorithms present a series of risks to media plurality in the UK describing a diverse media as “the cornerstone of a well-functioning democratic society.”

Ofcom cited the lack of transparency surrounding prominence and discoverability, reduced exposure to news sources, and the consumer’s ability to understand why algorithms serve them the news content they see online as major concerns.

In its response to Ofcom’s consultation, the News Media Association said: “There have been concerns that search engines operate on political biases but owing to the lack of transparency over algorithmic inputs and decision making this is unknowable.

“However, we do know that search engines rate recency over provenance, meaning that a news publisher who invested in breaking a story may not receive prominence for that content.

“Publishers who subsequently cover a story may receive more traffic. This creates a false incentive for publishers, who need traffic.”

Ofcom cited a lack of transparency surrounding algorithms as their first concern: “News publishers, industry bodies and academics considered that there is a lack of transparency as to how algorithms that serve news content decide on the prominence and discoverability afforded to different news sources and stories.”

“This issue presents a potential concern about the level of influence any one intermediary may be able to exercise over the range of viewpoints that citizens can access and consume, including where these might restrict the variety of viewpoints that citizens might be exposed to, and over the political agenda and political process.”

The Guardian Media Group argued that editorial decisions have been “redistributed throughout the news media value chain, including intermediaries’ algorithms” and News UK maintained that algorithms were “crucial to determining the prominence of news sources and what the user sees, and that they play an editorial function.”

DMGmedia also raised concerns over whether intermediaries’ commercial and political interests play a role in how algorithms serve news content to users.

The second concern Ofcom raised was that algorithms can also reduce exposure to a plurality of news sources and stories, with the statement noting: “several respondents suggested that international and national content is favoured over regional or local content, and that English language content is favoured over minority language content.”

Ofcom noted the work of Professor Dr Natali Helberger and colleagues for the Canadian government, which proposed “recommendation algorithms” which seek to “promote public values through exposure diversity” instead of optimising ‘clicks’.

Ofcom’s third concern related to media literacy in the UK and the consumer’s ability to recognise why algorithms served certain online news content to them, allowing them to critically engage with the news content they see.  

Ofcom also made a series of recommendations to the Secretary of State, advising that they retain the cross-media ownership and appointed news provider rules. Ofcom also recommended that the framework of the Media Public Interest Test should go “beyond print newspapers and broadcasters to capture a broader range of “news creators”.

“We think that this would be more reflective of the way in which people access and consume news today,” the regulator said in the report. 

Ofcom will now work to establish whether and how these issues present concerns to maintaining media plurality in the UK and to consider options for addressing these concerns. They aim to present their findings by summer 2022 and will co-ordinate their work with the Competition and Markets Authority, the Digital Markets Unit and members of the Digital Regulation Cooperation Forum.