Reports from content creators highlight a significant flaw in Instagram's automated copyright detection system, with users claiming even public domain music is being erroneously flagged. This issue, described by one user as a "broken AI-driven system," casts doubt on the platform's content moderation accuracy and impacts creators attempting to share legally permissible material.
Ronen▼, a user on the platform, recently shared their frustration, stating, > "If you try to post public domain music from 100 years ago to instagram, even using their own app's music tool, it will get a copyright takedown. Broken AI-driven system." This direct account underscores a growing concern among the creative community regarding Instagram's automated content ID.
Instagram's official policy acknowledges that works in the "public domain" are freely available for use, as their copyright protection has expired. However, despite this clear guideline, numerous users, including independent artists, have reported instances where their original or public domain music is incorrectly identified as infringing copyrighted material. This suggests a disconnect between policy and the practical application of the platform's AI.
The automated content identification system, designed to protect copyright holders, appears prone to false positives. Discussions across online forums like Reddit reveal widespread frustration, with users frequently labeling Instagram's copyright system as "broken" and a "known issue." Some reports even indicate third parties falsely claiming ownership of licensed music, leading to illegitimate takedown requests that further complicate the system's reliability.
The implications extend beyond music, as Instagram's AI has also been criticized for incorrectly flagging images as AI-generated. While Meta continues to invest in AI for content moderation and even music generation, the current user experiences highlight persistent challenges in ensuring the accuracy and fairness of its automated detection mechanisms, leaving creators vulnerable to unwarranted content removal.