<\/p>\n
YouTube<\/a> will now automatically apply AI content labels to realistic AI-generated videos on its platform \u2013 even when creators have not disclosed the use of AI themselves.<\/p>\n For music<\/strong>, the implications are pointed: the prominent label will apply to photorealistic AI music videos<\/strong> but not to stylized or animated ones, creating an implicit incentive for artists experimenting with AI visuals to favor the latter.<\/p>\n The platform said it is rolling out \u201cnew internal signals\u201d to identify AI-generated content, starting in May 2026<\/strong>.<\/p>\n If a creator does not specify whether AI was used, but YouTube<\/strong>\u2018s systems detect \u201csignificant photorealistic AI use,\u201d the platform said it will \u201cautomatically apply a label.\u201d<\/p>\n The announcement follows a recent expansion of YouTube\u2019s AI likeness detection tool to all eligible creators aged 18 and over. The tool, which the company has compared to Content ID, scans for AI-generated content featuring creators\u2019 likenesses.<\/p>\n That tool was first launched in October 2025<\/strong> to a limited group of creators, before being extended to celebrities and talent agencies<\/a> \u2013 including CAA<\/a><\/strong>, UTA<\/strong>, WME<\/strong>, and Untitled Management<\/strong> \u2013 in April 2026<\/strong>.<\/p>\n The latest move marks a shift from YouTube<\/strong>\u2018s previous approach, which relied on creators to self-disclose when they used AI tools to produce realistic synthetic content \u2013 a requirement first announced in late 2023<\/a>.<\/p>\n \u201cWhile we still require creators to manually disclose when they use realistic AI, we want to make the process more seamless and reliable,\u201d the platform said.<\/p>\n Creators who believe their content has been incorrectly flagged will be able to update their disclosure status in YouTube Studio<\/strong>.<\/p>\n However, in certain cases, AI disclosures will be permanent and cannot be removed.<\/p>\n Those cases include content created using YouTube<\/strong>\u2018s own AI tools, such as Veo<\/strong> or Dream Screen,<\/strong>\u00a0and content containing C2PA<\/strong> metadata indicating it was fully generated by AI.<\/p>\n Alongside the automatic detection system, YouTube<\/strong> is also making its existing AI content labels more prominent.<\/p>\n For long-form videos, the disclosure label will now appear directly below the video player, above the description.<\/p>\n For Shorts<\/strong>, the label will appear as an overlay on the video itself.<\/p>\n YouTube<\/strong> said this is \u201cnow the single label format for all photorealistic and meaningfully AI altered or generated content on YouTube.\u201d<\/p>\n Content that is \u201cunrealistic, animated, or slightly altered\u201d will continue to carry its disclosure in the expanded description only.<\/p>\n \u201cThese changes are designed to balance transparency with creator control,\u201d the platform said.<\/p>\n \u201cIt\u2019s important to note that a disclosure label alone does not change how a video is recommended or whether it\u2019s eligible to earn money.\u201d<\/p>\n The updates follow comments from YouTube<\/strong>\u2018s Global Head of Music, Lyor Cohen<\/a><\/strong>, who said in a letter to music industry partners in March<\/a> that YouTube<\/strong> is \u201cdoubling down\u201d on systems like Content ID<\/strong> to \u201cbuild new guardrails for likeness detection\u201d while \u201ccombating the spread of low-quality AI content.\u201d<\/p>\n In the same letter, Cohen<\/strong> quoted YouTube<\/strong> CEO Neal Mohan<\/strong> as saying: \u201cAI will remain a tool for expression, not a replacement.\u201d<\/p>\n Mohan<\/strong> had earlier identified managing \u201cAI slop\u201d and detecting deepfakes as priorities in his annual letter in January 2026<\/strong>.<\/p>\n Elsewhere in the DSP landscape, in terms of music specifically, Deezer<\/a> claims to be the first streaming platform to independently detect and tag AI-generated music \u2013 a capability it has been developing since early 2025, when it first deployed its AI detection tool, and expanded in June 2025 with explicit tagging.<\/p>\n Deezer<\/strong> said in April<\/strong> that 75,000<\/strong> AI-generated tracks now flood the platform daily, representing 44%<\/strong> of all new music uploaded.<\/p>\n Apple Music<\/a><\/strong> and Spotify<\/a><\/strong> have taken a different approach, placing the onus on the supply chain rather than building platform-level detection.<\/p>\n Apple Music<\/strong> launched its Transparency Tags system in March<\/a>, asking labels and distributors to declare AI-generated content at the point of delivery.<\/p>\n Spotify<\/strong> began showing AI credits in Song Credits in April<\/a>, but the feature depends on voluntary disclosure by artists through their label or distributor.<\/p>\n Spotify<\/strong> itself acknowledged: \u201cBecause we depend on artist disclosure, the absence of a credit doesn\u2019t mean AI wasn\u2019t used.\u201d<\/p>\n The labeling push arrives at a time when the implications of AI disclosure for music are under growing scrutiny.<\/p>\n A peer-reviewed study published in March found that listeners engage less deeply with music labeled as AI-generated<\/a> \u2013 even when the music was actually human-composed.<\/p>\n