CASE STUDY
As music streaming platforms have grown more sophisticated, personalization and recommendations have become defining features of the user experience. Spotify, in particular, has built a large part of its brand around algorithmic recommendations examining two main components: content-based filtering and collaborative filtering to adapt to users’ listening habits, preferences, and behaviors (Pastukhov, 2025). In recent years, Spotify has “tried to make listening feel personal” through playlists that predict what users might enjoy and allowing users to shape and update their playlists with more precision (Söderström, 2025). But now, the platform is looking to use users’ words, ideas, and creativity to power their recommendations through more complex algorithms (Söderström, 2025). In 2018, Spotify filed a patent describing a plugin called “MoodDJ” that could use speech recognition to determine users’ emotional states, gender, age, or accent to recommend and curate playlists accordingly (Stassen, 2021). The patent was granted in 2021 and, although the technology has not yet been implemented, its very existence has sparked debate about the boundaries of data collection, privacy, and consent on digital platforms.
According to the patent, Spotify’s MoodDJ would extract “intonation, stress, rhythm, and the likes of units of speech” which could allow the app to define and categorize a user’s mood (Walsh, 2021). With this information, the Spotify app would then combine that data with a user’s listening history and past preferences to determine appropriate music recommendations for their mood (Walsh, 2021). Supporters argue that MoodDJ represents a natural evolution of recommendation systems, while critics warn that emotional data crosses a line into overly invasive surveillance. Since emotions are deeply personal and often situation-dependent, the use of such technology raises concerns not only about privacy, but also about how much control users have over how their data is interpreted and used.
Proponents of Spotify’s MoodDJ emphasize the potential benefits of emotion-based personalization. From this perspective, the ability to detect mood could allow Spotify to deliver more relevant recommendations to users. Music has long been used as a tool for emotional regulation, reducing stress and anxiety, and general mood enhancement since when individuals listen to music, it activates the areas of the brain associated with pleasure, reward, and emotional regulation (Pierce, 2023). Therefore, more accurate emotional recommendations could improve well-being, increase satisfaction, and deepen engagement with the platform. In one study, Spotify researchers built a playlist based on participant’s emotions to model the new technology and multiple users commented on the potential for mental health benefits, from users with psychological diagnoses to those just navigating everyday emotions (Lowe-Brown et al., 2025). Moreover, participants added that using science to create these emotion-regulation playlists could be an effective alternative to medicine, as an overwhelming 18 out of 22 participants reported they would use MoodDJ (Lowe-Brown et al., 2025).
Advocates also argue that emotional inference mirrors everyday human behavior. In face-to-face interactions, people constantly draw conclusions about others’ emotions, social status, or intentions based on tone of voice, facial expressions, context, body language, etc. From this standpoint, Spotify’s MoodDJ could be seen as a digital extension of normal social inferences, not an unprecedented invasion. If individuals can reasonably infer that someone is stressed, angry, or relaxed based on how they speak in public spaces, proponents argue that a platform analyzing voices and words during user interaction operate within similar ethical boundaries.
Critics, on the other hand, argue that this comparison overlooks key differences between human inference and digital surveillance. Unlike typical social observation, Spotify’s MoodDJ would operate constantly and have the ability to store that information, which would surely be monetized over time (Access Now, 2021). Emotional data, in particular, is considered highly sensitive, as it can reveal mental states, diagnoses, vulnerabilities, and/or personal circumstances that users may not intend to share. Inferring emotions from speech could also be inaccurate or biased, leading to users being targeted by the platform in varying ways (Lowe-Brown et al., 2025).
Privacy advocates also point to the broader risks associated with data security and misuse. Even if Spotify’s intentions are to help their users –as affirmed by their researchers who recognize that digital history is personal, sensitive, and must be treated with the appropriate consideration–emotional data could be valuable to advertisers, third parties, or even hackers if accessed improperly (Anderson et al., 2020 and Lowe-Brown et al., 2025). From an autonomy standpoint, having Spotify automatically selecting music based on inferred emotions may limit users’ control over their listening experience, especially if the algorithm incorrectly analyzes their mood or the emotion behind a song (Lowe-Brown et al., 2025).
While the emotional regulation was pointed out as a major potential benefit of this system, participants further worried about potential safety risks, such as becoming “addicted to MoodDJ for emotion regulation” or using it to ruminate and prolong negative feelings (Lowe-Brown et al., 2025). Such safety concerns further complicate the issue, since the emotional data could be misused, exploited, or exposed, potentially placing users in vulnerable situations (Lowe-Brown et al., 2025).
Access Now, Fight for the Future, Union of Musicians and Allied Workers, and a group of over 180 musicians and human rights groups from around the world have publicly opposed the potential implementation of Spotify’s MoodDJ in a letter from April 2, 2021 (Access Now, 2021). These groups express concern that algorithmically curating music based on inferred emotions brings about major concerns regarding emotional manipulation, discrimination, privacy violations, data security, and exacerbating inequalities in the music industry. From their perspective, emotion-based curation risks prioritizing algorithmic efficiency and profits over the human, relational, and emotional nature of music consumption. On April 15, 2021, Spotify replied to the letter, stating that the company “has never implemented the technology described in the patent in any of [its] products and ha[s] no plans to do so.” (Access Now, 2021). Nonetheless, Access Now continues to call on Spotify to publicly commit to never “use, license, sell, or monetize the technology” (Access Now, 2021).
At its core, the debate surrounding Spotify’s MoodDJ highlights the broader ethical tensions between innovation and surveillance. While supporters emphasize the potential for improved emotional regulation and well-being in additional to music discovery, critics raise concerns about autonomy, safety, data privacy, and fairness. Since emotional data is uniquely sensitive and deeply personal, the acceptability of such technology depends on not just what is technically possible, but also on how transparently it is implemented, used, and governed.
DISCUSSION QUESTIONS
- What ethical concerns arise when a platform uses speech and situational data to infer users’ emotional states for personalized content recommendations?
- How does emotion-based music curation challenge user autonomy, especially when algorithms may misinterpret emotions or associate music with unwanted memories or moods?
- How does Spotify’s proposed technology impact artists, particularly in terms of artist intent, emotional interpretation, and algorithmic control over how their music is presented to audiences?
- How does Spotify’s technology differ from people normally inferring emotions in everyday interactions and why do these differences, if any, matter?
- What safeguards, transparency guidelines, or consent questions would be necessary for emotion-based personalization to be ethically acceptable, if at all?
ADDITIONAL INFORMATION
- Access Now. (2021, April 15). Dear Spotify: don’t manipulate our emotions for profit. Available at: https://www.accessnow.org/press-release/spotify-tech-emotion-manipulation/#:~:text=Emotion%20manipulation:%20Monitoring%20emotional%20 state,not%20covert%20manipulation%20and%20monitoring.%E2%80%9D
- Anderson, , Gil, S., Gibson, C., Wolf, S., Shapiro, W., Semerci, O., & Greenberg, D. M. (2020, December 15). ‘Just The Way You Are:’ Music Listening and Personality. Spotify Research. Available at: https://research.atspotify.com/2020/12/just-the-way-you-are-music-listening-and-personality
- Lowe-Brown, Z., Glasser, S., Koval, P., & Wadley, G. (2025, September 29). Proceedings of the 2024 CHI Conference on Human Factors in ACM Digital Library. Available at: https://dl.acm.org/doi/10.1145/3726986.3726998
- Pastukhov, D. (2025, September 1). Inside Spotify’s Recommendation System: A Complete Guide (2025 Update). Music Tomorrow. Available at: https://www.music-com/blog/how-spotify-recommendation-system-works-complete-guide
- Pierce, (2023, December 19). The Science Of Music And Emotional Regulation: Choosing The Right Soundtrack. Life Skills Advocate. Available at: https://lifeskillsadvocate.com/blog/music-and-emotional-regulation/
- Söderström, G. (2025, December 10). You’re in Control: Spotify Lets You Steer the Algorithm. Spotify Newsroom. Available at: https://newsroom.spotify.com/2025-12-10/spotify-prompted-playlists-algorithm-gustav-soderstrom/
- Stassen, M. (2021, January 27). Spotify’s latest invention monitors your speech, determines your emotional state… and suggests music based on it. Music Business Worldwide. Available at: https://www.musicbusinessworldwide.com/spotifys-latest-invention-will-determine-your-emotional-state-from-your-speech-and-suggest-music-based-on-it/
- Walsh, (2021, January 30). Spotify patents tech to recommend songs based on users’ speech, emotion. Axios. Available at: https://www.axios.com/2021/01/30/ spotify-patent-users-speech-recommend-music
ACKNOWLEDGMENTS
This case was supported by funding from the John S. and James L. Knight Foundation. It can be used in unmodified PDF form in classroom or educational settings. For use in publications such as textbooks, readers, and other works, please contact the Center for Media Engagement at mediaengagement@austin.utexas.edu.
SUGGESTED CITATION
Mohan, M., Williams, K., & Stroud, S. R. (September 2026). Soundtracking Surveillance? The Ethics of Spotify’s MoodDJ. Media Ethics Initiative, Center for Media Engagement. https://mediaengagement.org/research/soundtracking-surveillance-spotify
