YouTube serves AI‑generated explicit songs to minors, NDTV finds

Scale of the problem on India’s massive YouTube audience

Researchers identified about two dozen channels and YouTube-created topics featuring AI-generated explicit songs, mostly reached through platform recommendations rather than direct searches. None of the documented songs carried age restrictions, and most lacked an "Explicit" tag.

The videos displayed typical AI traits, including repetitive high-pitched music, inconsistent lyrical flow, polished thumbnails and no conventional music videos.

Mimansa Singh Tanwar warned that exposure to this explicit content can lead to behavioral issues, normalize aggressive sexual behavior and objectify gender attitudes in developing children.

Detection tools flag AI‑made tracks and YouTube’s labeling

NDTV tested 16 songs and found all were likely AI-generated using Quicksilver and SubmitHub detection tools. One channel created in January 2026 uploaded 177 songs by 28 October 2026, with 153 of those posted in the preceding 30 days. This activity averaged roughly five songs per day.

Deezer, the streaming platform from France, stated that in June 2026 AI‑created songs were uploaded at an average rate of roughly 90,000 per day. Most channels involved did not credit identifiable singers, lyricists or composers and reused the same artist names across multiple channels. These channels also lacked verifiable social media accounts for the supposed artists.

YouTube says it determines whether to prohibit sexually explicit content by considering if language is graphic or lewd. When applying these rules, the platform also assesses if sexual material is sustained rather than brief.

Experts warn of mental‑health and aggression risks

Although NDTV’s catalog of songs was entirely sexual from beginning to end, YouTube placed no age‑restriction on any of them. Some songs appeared to reference schoolchildren, and these tracks remained available on the platform despite YouTube policies prohibiting sexual content involving minors.

Penalties for publishing or sending obscene content electronically are stipulated in Section 67 of the Indian Information Technology Act. A first‑time offence under this statute can result in a jail term of up to three years.

Indian law and safe‑harbour rules apply to obscene AI content

While user‑generated content on online services is shielded by safe‑harbour provisions, such immunity hinges on meeting due‑diligence requirements.

The Ministry of Electronics and Information Technology, on December 29 2025, cautioned intermediaries about these duties and warned that neglecting them could endanger their safe‑harbour safeguards.

New amendments covering AI‑generated and synthetic media have added further compliance duties for intermediaries handling unlawful synthetic content, such as obscene or sexually explicit material.

YouTube recommendations push explicit AI songs to users

Advocate Aditya Joshi of Delhi noted that US‑based intermediaries like YouTube frequently seem unwilling to delete material solely based on complaints. He further explained that their moderation tools confront two key hurdles: spotting troublesome keywords in regional languages and enhancing AI detection of harmful material.

Based in San Bruno, California, YouTube is an American online platform for sharing videos.

Future enforcement and compliance pressures on YouTube

In February 2005, three ex‑PayPal staff members—Chad Hurley, Steve Chen, and Jawed Karim—established YouTube. Google bought YouTube in November 2006 for $1.65 billion, and it now operates as one of Google’s subsidiaries.