Europe’s future rules for online platforms should apply to all companies, not just the big ones, a senior YouTube executive told POLITICO.
The European Commission is mulling whether to impose “enhanced responsibilities” for very large platforms as part of its upcoming rulebook called the Digital Services Act, but the video-sharing platforms Chief Product Officer Neal Mohan warned against it.
“Anything that the DSA does has to apply to all companies, all platforms, not just the large ones but the small ones as well. Platforms that are small today can become large in a matter of months,” he said.
The legislation, expected on December 9, will update Europe’s decades-old framework for the online world by defining content moderation rules for platforms including YouTube, which is owned by Google. The DSA could set transparency obligations on how platforms recommend content to users, how they decide which material to remove and how online advertising works.
Tech companies such as Facebook, Twitter, YouTube and most recently Chinese-owned TikTok, have faced immense pressure in recent years to police content on their platforms as hate speech, disinformation surrounding elections and conspiracy theories like QAnon continue to proliferate.
Depending on how far-reaching the Digital Services Act will be, YouTube could be forced to provide much more information to regulators, researchers and, in some cases, the general public on how the company operates.
YouTube has its own list of demands to Brussels on what upcoming legislation must provide: “Clarity around notice and take down, with a specific focus on illegal content; flexibility around recommendations; and the importance of regulation applying in a universal fashion,” Mohan said.
One regime to rule them all
The Commission is considering specific provisions — known as asymmetric rules — “for very large platforms to mitigate systemic risks.” In the Parliament, the center-right European People’s Party recently backed asymmetric rules, arguing that the largest platforms should have to comply with more rules than startups and small businesses.
YouTube is squarely in the “very large” category. It has two billion monthly users across the globe and 500 hours of content are uploaded every minute.
Mohan argued that with rules requiring only the biggest platforms to police content, such material would just wind up elsewhere. “The type of content they may be looking to remove from large platforms will then show up on smaller platforms and still expose users throughout the EU to that type of content,” he said, arguing that users move “very easily” from one platform to the other.
That is happening in the U.S. Millions of conservative users have recently left Facebook and Twitter to join alternative social media platforms such as Parler, which has fewer rules about flagging misinformation about the U.S. election’s outcome.
But Jan Penfrat, a senior policy advisor at digital rights NGO EDRi, said asymetric rules would still be a huge step in curbing illegal and harmful posts. “It makes sense to have a specific regime for the largest platforms because the problem we see with illegal and harmful content is worse on them. You can post hateful comments on a small platform, if nobody sees them, then there is less of an issue,” he said.
“It’s totally in YouTube’s interest to ask for similar rules for everyone because it’s very easy for them and harder for smaller competitors to apply them — it would cement their position,” Penfrat added.
Illegal vs. legal
Regulators have struggled limiting the spread of disinformation because much of it is technically legal. Mohan said that given the “subjective” nature of legal content that’s still harmful, platforms themselves are best equipped to decide how to police that content instead of regulators.
“The nature of potentially harmful content changes on a regular basis. Instead of trying to capture that in a statutory regime, why don’t platforms have the flexibility to enforce that within their own community guidelines in a broad sense?” he said.
In the Parliament, the European People’s Party unsuccessfully campaigned to include harmful content in the MEPs non-binding reports. And EU countries themselves are divided on the issue.
Sweden wants the DSA to focus on specifically on illegal content, but France is a big proponent of more expansive legislation. Paris recently argued that it is precisely because platforms such as YouTube, Facebook and TikTok already remove legal yet harmful content if they go against their own terms of service that such practices should also be regulated and supervised.
Still, French authorities acknowledge that when it comes to disinformation, measures to reduce the content’s virality — downgrading its visibility or limiting sharing — would be more appropriate than removals and blocking.
Why do we see what we see?
Over the years, YouTube’s recommendation algorithm has been accused of fueling radicalization by steering users towards inflammatory material.
Policymakers, especially in the European Parliament, would like to better understand how recommandations work and to what extent the algorithm is fed by user behavior and personal data. A non-binding report on the DSA even suggested allowing users to opt out of allowing platforms to suggest content, a mechanism known as content curation.
Asked whether YouTube would be ready to make its recommendation algorithm more transparent, Mohan said: “What we have tried to focus on is to give users control on how they can manage their recommendations, filter out pieces of content.”
“Our users have told us repeatedly that the recommendations that YouTube provides on the type of content that shows up in their home feed is a critical component to how they use the product. Algorithms are core to the way users experience YouTube, it’s what put them in the driver’s seat,” Mohan argued.
But regulators take a different view. Commission Executive Vice President Margrethe Vestager said the opposite in October, arguing that platforms had the power to determine what citizens see online. They are “guardians that decide what we can and can’t see,” she said. The DSA’s goal, then, is to “make sure that we, as a society, are in control.”
The legislation will require platforms to inform users how the recommender systems decide which content to show and give them the tools to “influence the choices that recommender systems make on our behalf,” she said. It is unclear what rules the Commission will propose to achieve that.
This article is part of POLITICO’s premium Tech policy coverage: Pro Technology. Our expert journalism and suite of policy intelligence tools allow you to seamlessly search, track and understand the developments and stakeholders shaping EU Tech policy and driving decisions impacting your industry. Email [email protected] with the code ‘TECH’ for a complimentary trial.