Submitted by Eric Shadowens, European Policy Lead
Patreon is grateful for the opportunity to submit evidence on the Draft Online Safety Bill. Patreon is a membership platform that empowers creators and artists to earn sustainable income. The platform, which was started in 2013 by musician and video creator Jack Conte and his college roommate Sam Yam, has become a top income-generating solution for over 200,000 creators. To date creators have earned over £1.5 billion pounds ($2 billion US dollars) through Patreon’s subscription-style payment model, including over £45 million this year in the UK. Within this evidentiary submission, we will focus on what we believe to be the best way to continue serving creators while also keeping them and their communities safe.
While Patreon certainly supports the noble goals of the proposed legislation, we will discuss the potential negative implications for creators and explain how the lack of clarity around important definitions within the bill may do more harm than good. Contradictory ideas within the legislation around such things as freedom of expression and requiring companies to remove “harmful” but not illegal speech, may make it much more difficult for digital businesses to operate with any certainty and result in that same impact on the many creators in the UK who rely on Patreon to earn a living. In addition, the compliance costs faced by companies like Patreon could be outsized compared to larger companies in the same industry.
The proposed law as currently defined does not create clear rules around which companies qualify as Category 1 user-to-user services, leaves the door open for political intervention in the regulatory process, and does not address how different platform content distribution models might affect the application of the rules. For instance, Patreon’s Community Guidelines explicitly state that because creators earn funds through our platform, we may consider “what (they) do with (their) membership off platform” in reviewing their accounts. This holistic approach is more time- and labor-intensive than the single-piece-of-content review approach of large distributed platforms such as Facebook or Twitter; rules that fail to acknowledge both approaches may disadvantage companies like Patreon. As the Committee considers requirements related to content reporting and review systems, and the potentially financially onerous obligations of this legislation for small- and medium-sized companies, it is important to account for the differences amongst companies within the digital space and avoid “one size fits all” solutions.
Patreon is committed to building a safe and supportive environment for creators in which they can grow their businesses and engage with their patrons. As more and more creators in the UK and elsewhere come onto the platform, it is crucial that we are able to provide them certainty as to what is required for them to maintain an account. We will focus on how the vague nature, and at times contradictory ideas, of the proposed law will make this more difficult and negatively impact creators.
The idea of “duty of care” itself is not necessarily a bad one. Certainly Patreon agrees that, as a company, we have an obligation to our community of creators and patrons to maintain a safe environment. This is expressed in our Community Guidelines, which in many cases, already goes further than what is legally compulsory. The key issue related to “duty of care” is that these requirements are subjective. This is clearly highlighted by potential requirements within the law for a company to take action against content that is deemed harmful, but crucially not illegal, for adults and/or children. The idea of harm is largely undefined and fails to reconcile the concerns related to freedom of expression. For instance, the law requires for the removal of content that is a “material risk of the content having, or indirectly having, a significant adverse physical or psychological impact on a [child/adult] with ordinary sensibilities.” This could include a wide variety of content. It certainly makes it very difficult to determine the difference between what someone may deem offensive and actually harmful, let alone the considerations of how you prove a company should have known about the potential for “indirect” psychological harm from one piece of content to the next.
Making companies the arbiters of what speech actually causes harm could lead to undue censorship at the users’ expense. The high costs of penalties for lack of compliance in this space, in addition to the ambiguous legal middle ground this legislation creates, puts companies in a position where removing content is always the safer option anytime there is doubt. It also begs the question as to why the government has not chosen to legislate further on what types of speech are illegal, especially if it believes that speech is inherently harmful. Certainly a model, like the EU’s Digital Services Act, at least offers more clarity and certainty to the conversation by focusing specifically on illegal content.
The above focus on potentially harmful speech contradicts the language in the legislation regarding “journalistic content” and content considered to be of “democratic importance.” These concepts are also given very broad definitions that will make compliance extremely difficult. If journalistic content is defined only as “content generated for the purposes of journalism,” how should platforms treat content by activists and extremists who claim to be journalists? Do they then have protection for content that is otherwise violating? A similar issue arises when discussing content that may be of “democratic importance,” which is only defined as “intended to contribute to democratic political debate.” For example, if hate speech is used as part of an argument to propose immigration restrictions, is that harmful or of democratic importance according to this law? The ambiguous nature of the regulation in this space only provides further uncertainty and confusion as to what content is intended to be within scope.
The proposed law suggests that there will be a higher burden placed on the largest companies which will ultimately be included in Category 1, though who is included is still unknown. This is a key provision that will determine whether or not this law will stifle innovation and competition in the digital space. The potential for onerous compliance costs, especially for smaller businesses, could solidify the largest companies market position while causing small and medium size businesses to assess the value of compliance versus their own presence in the market. It is crucial that the Committee take this into account and ensure the bill ultimately does not punish a company for growing.
The Committee must also ensure that the legislation treats different content models differently. As mentioned previously, distributed content models place an onus on identifying and reviewing a huge amount of individual pieces of content from millions, if not billions, of different users. While Patreon certainly does consider content posted on our site against our policies and has the mechanisms in place to review that content, including both technical solutions and manual review, our focus is as much on what account level action that may merit. Certainly use of harmful content such as terrorist content or obvious hate speech, even on another platform, may result in removal. However, a decision to close down a creator’s access to their business is something we at Patreon take very seriously and only do so once we have considered the full context of the case. This is why we have a reporting flow that allows the reporter to provide more context, such as links to external sites, to help in our review. This is particularly important and also displays a very real difference between our review considerations and those at larger platforms with different content distribution models. If the legislation moves forward with required changes to our reporting flow, forced implementation of different algorithmic censors, etc… it may divert important resources away from this full service review work without actually solving for the real issues we face as a platform.
Patreon is committed to making the internet safer and ensuring the empowerment of a diverse community of voices. Achieving that goal will require certainty and objectivity, however; the Online Safety Bill in its current form exposes companies like Patreon and our creators in the UK, to significant risk. We believe the bill needs to be clear and codify exactly what content it wants to regulate rather than leaving so much up for interpretation. We believe the bill needs to clearly define what content it intends to regulate, and to empower the regulator with autonomy and freedom from influence by political stakeholders. Finally, the Committee must reconcile how the compliance costs of this bill may directly limit competition in the market to avoid a situation where the only companies that can afford compliance are those that already have dominant market positions.
Patreon applauds the Committee’s commitment to finding consumer-first solutions for these challenging subjects. UK-based creators stand to benefit from more clearly defined expanded protections that consider the nuances of different business models that promote digital innovation and competition. We are grateful for the Committee’s consideration of our perspective and are willing to provide additional background, information, and insights into this matter as needed.
22 September 2021