Instagram Confirms Algorithm Glitch Over-Promoted Creed's 'Higher'

Key points

• Instagram confirmed an algorithmic error caused Creed's 'Higher' to be over-recommended to users.
• The platform's engineering team is currently investigating and fixing the issue.
• Users reported the song appearing for unrelated content, such as food and pets, sparking conspiracy theories.
• Instagram clarified that recommendations for Soulja Boy's 'Turn My Swag On' are not a bug but reflect actual popularity.
• The glitch occurred despite the algorithm typically weighing user history and visual content for song suggestions.

Background

This incident follows recent discussions about Instagram's algorithmic transparency, including CEO Adam Mosseri's earlier warnings that removing algorithms would flood users with brand content. The current glitch highlights the complexities of automated content curation on the platform.

How outlets are covering it

KSL, Rolling Stone, and Consequence of Sound all report that Instagram confirmed the Creed recommendation was an unintended algorithmic error. Rolling Stone and Consequence of Sound emphasize the user confusion and the subsequent apology from Instagram. KSL notes the user speculation about a conspiracy, while all sources agree that the Soulja Boy recommendation is a separate, intentional feature based on popularity. No source disputes the core facts, but they differ in emphasis: Rolling Stone focuses on the user experience and the apology, while KSL highlights the conspiracy theories and the engineering team's response.

Why it matters

The incident underscores the potential for algorithmic errors to create widespread user confusion and misinformation. It also highlights the challenges platforms face in balancing personalized recommendations with accurate content curation, potentially affecting user trust in automated systems.

What to watch

Instagram's engineering team is expected to resolve the Creed recommendation glitch. The platform may also review its algorithmic processes to prevent similar errors in the future. Users can expect normal recommendation patterns to resume once the issue is fixed.