Executive summary

Twitch is a live streaming service built around interactive, real-time communities, where creators engage in a wide variety of different activities, including video games, art, cooking, and other types of creative content.

At Twitch, we are committed to fostering a dynamic and inclusive environment that enables streamers to express themselves safely while ensuring a positive, engaging environment for viewers, free of illegal and harmful interactions. This starts with Twitch’s Community Guidelines, which seek to balance user expression with community safety, and set the rules for the behaviour of everyone on Twitch. Developed in consultation with external safety, human rights, and policy experts, these guidelines are regularly reviewed and updated to reflect and respond to the community’s evolving needs.

We identify and address potential safety risks using a combination of automated detection, proactive human review, and user reporting. Our global Trust and Safety operations team works to quickly assess content and accounts flagged by users and by our automated detection models. The speed at which we respond and take action on user reports is critical given the live nature of Twitch, and in 2024, we responded to 82.6% of reports in under 10 minutes and 93.4% of reports in under an hour. Twitch employs extensive  human review to help ensure that any enforcement action we take remains accurate and fair for our community members. 

Twitch recognizes the risks posed by misinformation, and we believe that individuals who use online services to spread false, harmful information do not have a place in our community. We maintain a dedicated policy to address misinformation: our Misinformation Actor policy. This policy targets individuals who repeatedly and persistently share , widely debunked misinformation that has the potential to cause real-world harm. These actors share three characteristics: their online presence—whether on or off Twitch—is dedicated to (1) persistently sharing (2) widely disproven and broadly shared (3) harmful misinformation topics, such as conspiracies that promote violence. We prohibit harmful misinformation actors who meet all three of these criteria since taken together they create the highest risk of harm, including the incitement of real-world harm. 

Our enforcement numbers under this policy are relatively low due to several factors. (i) The mechanics of Twitch are not conducive to spreading misinformation or investing in large-scale disinformation campaigns. It is extremely difficult for a new streamer to garner large numbers of concurrent viewers; it takes time to grow an audience on Twitch. Most Twitch content is also long-form and ephemeral.  Instead, once a livestream ends, the content typically disappears or is only available in limited formats.  The ephemeral nature of content on Twitch makes large-scale misinformation campaigns much harder to execute compared to other UGC video and social media services where content remains indefinitely and can be amplified more easily. (ii) Our targeted policy applies to those who persistently share harmful misinformation. Due to the long-form nature of Twitch’s content, we are focused on a streamer’s aggregated content rather than a specific, isolated statement within a longer piece of content. (iii) When we launched our Harmful Misinformation Actor policy, we took swift action against accounts that posed a threat to our community. We believe enforcement of our policy—particularly upon its adoption in 2022—has been an effective deterrent to harmful misinformation actors; we have not seen large numbers of them attempt to join our service.

Even if someone is not a Harmful Misinformation Actor, Twitch takes action against  misinformation that targets specific communities under our Hateful Conduct & Harassment policies, and we take action on content that encourages others to engage in physically harmful behaviour under our Self-Destructive Behaviour policy.

In addition to misinformation, Twitch invests significant resources to ban bots, spammers, impersonators, and other types of bad actors to combat inauthenticity on our service. We have automated and proactive detection systems that work in tandem with our reporting system to identify and remove bots, known bad actors, and those who are trying to evade a suspension or ban.

While misinformation is not currently prevalent on Twitch, we recognize the harm that this content can cause, particularly when it is related to an election. At the start of 2024, we assembled a cross-functional working group—across the Product, Policy, Operations, Legal, Risk Management, and Content teams—to ensure Twitch was prepared and could stay ahead of potential harms related to EU and US elections. Our Trust & Safety operations team did not observe any misinformation-related—or hateful conduct, harassment, or violence-related—threats around the EU election or any election that took place in EU Member States during the past year.

We continuously refine our approach to safety, drawing on insights from experts and evolving trends in our community. Recognizing that the  prevalence of harmful misinformation can shift, we actively collaborate with industry, academia, and civil society to assess emerging risks and adjust  our strategies accordingly. To reinforce our commitment to fighting misinformation, we became a signatory to the Australian Code of Practice on Disinformation and Misinformation at the end of 2023, fostering stronger cross-sector collaboration and information sharing. Additionally, we participate in a variety of global industry knowledge-sharing initiatives, including the New Zealand Code of Practice for Online Safety and Harms (which also addresses disinformation), the EU Hate Speech Code, the EU Internet Forum, and the Global Internet Forum to Counter Terrorism (GIFCT).

The EU Code of Practice on Disinformation serves as a valuable mechanism for information sharing and collaboration that will help strengthen industry’s abilities to react quickly to the spread of misinformation. As a signatory, Twitch’s goal is to make meaningful contributions and learn from expert organisations and industry peers. We are committed to combating misinformation on Twitch in an effective yet targeted manner that balances freedom of expression with keeping our communities safe.