\n<\/p>\n
Elon Musk\u2019s X is the latest social network to roll out a feature to label edited images as \u201cmanipulated media,\u201d if a post by Elon Musk is to be believed. But the company has not clarified how it will make this determination, or whether it includes images that have been edited using traditional tools, like Adobe\u2019s Photoshop.<\/p>\n
So far, the only details on the new feature come from a cryptic X post from Elon Musk<\/a> saying, \u201cEdited visuals warning,\u201d as he reshares an announcement of a new X feature made by the anonymous<\/a> X account DogeDesigner<\/a>. That account is often used as a proxy for introducing new X features, as Musk will repost from it to share news.<\/p>\n Still, details on the new system are thin. DogeDesigner\u2019s post claimed X\u2019s new feature could make it \u201charder for legacy media groups to spread misleading clips or pictures.\u201d It also claimed the feature is new to X.<\/p>\n Before it was acquired and renamed as X, the company known as Twitter had labeled tweets<\/a> using manipulated, deceptively altered, or fabricated media as an alternative to removing them. Its policy wasn\u2019t limited to AI but included things like \u201cselected editing or cropping or slowing down or overdubbing, or manipulation of subtitles,\u201d the site integrity head, Yoel Roth, said in 2020.<\/p>\n It\u2019s unclear if X is adopting the same rules or has made any significant changes to tackle AI. Its help documentation<\/a> currently says there\u2019s a policy against sharing inauthentic media, but it\u2019s rarely enforced, as the recent deepfake debacle<\/a> of users sharing non-consensual nude images showed. In addition, even the White House now shares manipulated images.<\/a><\/p>\n Calling something \u201cmanipulated media\u201d or an \u201cAI image\u201d can be nuanced. <\/p>\n Given that X is a playground for political<\/a> propaganda<\/a>, both domestically and abroad, some understanding of how the company determines what\u2019s \u201cedited,\u201d or perhaps AI-generated or AI-manipulated, should be documented. In addition, users should know whether or not there\u2019s any sort of dispute process beyond X\u2019s crowdsourced Community Notes.<\/p>\n Techcrunch event<\/p>\n \n\t\t\t\t\t\t\t\t\tSan Francisco<\/span> As Meta discovered when it introduced AI image labeling in 2024, it\u2019s easy for detection systems to go awry. In its case, Meta was found to be incorrectly tagging<\/a> real photographs with its \u201cMade with AI\u201d label, even though they had not been created using generative AI.<\/p>\n This happened because AI features are increasingly being integrated<\/a> into creative tools used by photographers and graphic artists. (Apple\u2019s new Creator Studio suite, launching today<\/a>, is one recent example.) <\/p>\n As it turned out, this confused Meta\u2019s identification tools. For instance, Adobe\u2019s cropping tool was flattening images before saving them as a JPEG, triggering Meta\u2019s AI detector. In another example,<\/a> Adobe\u2019s Generative AI Fill, which is used to remove objects \u2014 like wrinkles in a shirt, or an unwanted reflection \u2014 was also causing images to be labeled as \u201cMade with AI,\u201d when they were only edited with AI tools.<\/p>\n Ultimately, Meta updated its label to say \u201cAI info,<\/a>\u201d so as not to outright label images as \u201cMade with AI\u201d when they had not been. <\/p>\n Today, there\u2019s a standards-setting body for verifying the authenticity and content provenance for digital content, known as the C2PA<\/a> (Coalition for Content Provenance and Authenticity). There are also related initiatives like CAI<\/a>, or Content Authenticity Initiative, and Project Origin<\/a>, focused on adding tamper-evident provenance metadata to media content. <\/p>\n Presumably, X\u2019s implementation would abide by some sort of known process for identifying AI content, but X\u2019s owner, Elon Musk, didn\u2019t say what that is. Nor did he clarify whether he\u2019s talking specifically about AI images, or just anything that\u2019s not the photo being uploaded to X directly from your smartphone\u2019s camera. It\u2019s even unclear whether the feature is brand-new, as DogeDesigner claims.<\/p>\n X isn\u2019t the only outlet grappling with manipulated media. In addition to Meta, TikTok<\/a> has also been labeling AI content. Streaming services like Deezer<\/a> and Spotify<\/a> are also scaling initiatives to identify and label AI music, as well. Google Photos is using C2PA<\/a> to indicate how photos on its platform were made. Microsoft, the BBC, Adobe, Arm, Intel, Sony, OpenAI, and others are on the C2PA\u2019s steering committee<\/a>, while many more companies have joined as members.<\/p>\n X is not currently listed among the members<\/a>, though we\u2019ve reached out to C2PA to see if that recently changed. X doesn\u2019t typically respond to requests for comment, but we asked anyway.<\/p>\n<\/div>\n
\n\t\t\t\t\t\t\t\t\t\t\t\t\t|<\/span>
\n\t\t\t\t\t\t\t\t\t\t\t\t\tOctober 13-15, 2026<\/span>\n\t\t\t\t\t\t\t<\/p>\n<\/p><\/div>\n<\/p><\/div>\n<\/div>\n