CAT 2026 Slot 2 VARC: Reading Comprehension Passage 3: Reading Comprehension – Streaming Algorithms and Musical Bias with Solutions

In my book "Screens", I document how major streaming platforms have exploited human creativity for their profit. We allow this to happen, I argue, because we also gain some entertainment from using their services. It is a dynamic that makes us complicit in these platforms' accumulation of wealth and data: we are simultaneously victims and beneficiaries. I describe this complicity, but I also enact it, through my own digital footprint: my Netflix watch history, my Spotify playlists, and yes, my Spotify Wrapped summaries.

People often describe streaming algorithms' curated playlists as "bland" or "homogenous"—the digital equivalent of elevator music. Spotify's systems are built to "sound like a friend," as their marketing puts it, recommending tracks that, if suggested by a person, would seem "cool," "empathetic," "energetic," and "relatable." The company describes these strategies as helping its service seem "personalized" and "effortless." This appears to be bound up with making us feel understood.

Trust is a challenge for streaming companies, partly because their algorithms regularly reinforce sexist, racist, and Western-centric patterns in music recommendations. While the platforms are working on these issues, they persist: a recent audit found that their latest recommendation engines generate demographic skews at a higher rate than previous versions. In the book, I wrote about the algorithmic biases and also demonstrated them with the apps. When I asked a popular playlist generator to produce a workout mix, it gave me an entirely male hip-hop lineup; when my friend used Spotify to find tracks similar to an acoustic folk song, the recommendations leaned toward generic pop country. Those weren't flukes. Industry data suggests both tendencies are widespread.

In my own Spotify sessions, I wanted to demonstrate how the platform's veneer of musical neutrality could lull us into absorbing biased or homogenous recommendations without much critical engagement. Over time, Spotify seemed to be nudging me toward mainstream pop—including suggesting I add a "trending" playlist featuring Ed Sheeran to my daily mix. I am not aware of research on whether Spotify tends to favor certain major labels or artists, and I can only guess why it seemed that way in my session. Spotify explicitly states that its algorithms shouldn't attempt to manipulate user tastes. When I questioned Spotify about some of these issues, it blamed biases in its collaborative filtering data—though I suspect my own arguably skewed listening history played a role too. The platform's explanation of its recommendation logic stated: "The recommendation system is designed to support listener satisfaction and engagement, although its suggestions can be helpful or misleading."

Spotify has its own goals, of course. Among them, it emphasizes wanting to build a platform that "connects artists and listeners globally." But even while the company promotes that mission, its shareholders seek a return on their investment. That will presumably require getting people subscribing to premium services even more than they already are—a goal that is easier to accomplish if people see those algorithms as trustworthy tastemakers.