AI labels and icons
The sparkle icon, content credentials and "Made with AI" labels, and what each one says about an image or feature.
These five pages cover the icons and labels that say artificial intelligence was involved: the sparkle icon ✨, the AI icon, the Content Credentials (CR) pin, Meta’s “AI info” label and the “Made with AI” label. They are of two kinds. The sparkle marks a button or tool in an app that uses AI; the other four label an image, video or text that AI made or changed. Each platform’s help centre explains its own label. These pages set the labels side by side and say what puts each one on a post, as the platforms and the European Commission described them in September 2026.
How the sparkle came to mean AI
The Nielsen Norman Group wrote in September 2024 that, with no standardised way to show a “mysterious AI-driven process”, the sparkles icon had become “the de facto symbol for everything AI”. In its research, not one participant mentioned AI when shown a sparkles icon. TechCrunch reported Twitter’s “sparkle button”, which switched between the algorithmic timeline and the latest tweets, rolling out to all iOS users on 18 December 2018. By the end of 2024 The Globe and Mail counted Google, Microsoft, Spotify, Samsung, Slack, LinkedIn and ChatGPT among the companies using a four-pointed star for AI features. The emoji behind it, ✨ SPARKLES, was new in Unicode 6.0 in October 2010.
Labels on content
The European Commission published a Code of Practice on marking and labelling AI-generated content on 10 June 2026, with a set of EU icons: a basic icon, and variants for content fully AI-generated and partially AI-modified. The Commission says using the icons is optional, but the labelling requirements of Article 50 of the AI Act are not; those apply from 2 August 2026. Meta began tagging photos “Made with AI” in May 2024, and renamed the label “AI info” in July after photographers complained, TechCrunch reported, that it was going on real photos with only basic edits. YouTube uses “Made with AI” in a video’s description for content that has been meaningfully altered or synthetically generated.
What the CR pin records
The Coalition for Content Provenance and Authenticity (C2PA) introduced the pin on 10 October 2023. It shows that a file carries Content Credentials, which can say where and when it was created, what tools were used, and whether generative AI was used. The C2PA says the credentials make no value judgment about whether content is true, and that the metadata can be removed. Platforms read the same data for their own labels: TikTok began reading Content Credentials in May 2024, and Meta reads the “AI generated” information in the C2PA and IPTC standards.
Sources
- The Proliferation and Problem of the ✨ Sparkles ✨ Icon, Kate Kaplan, 20 September 2024, Nielsen Norman Group, Sparkles Are Ambiguous; Sparkles as “All Things Artificial Intelligence”
- Twitter rolls out 'sparkle button' to let users hide the algorithmic feed, Lucas Matney, 18 December 2018, TechCrunch, article
- Why the sparkle emoji became the symbol of AI, Benjamin Errett, 28 December 2024, The Globe and Mail, article
- DerivedAge.txt, Unicode Character Database, Unicode Consortium, Unicode 6.0.0 section
- Commission publishes Code of Practice on marking and labelling AI-generated content, Press release, 10 June 2026, European Commission, press release
- EU Icons for labelling AI-generated content, European Commission, Shaping Europe’s digital future (page updated 24 September 2026), European Commission, whole page
- Meta changes its label from 'Made with AI' to 'AI info' to indicate use of AI in photos, Ivan Mehta, 1 July 2024, TechCrunch, article
- Our Approach to Labeling AI-Generated Content and Manipulated Media, Monika Bickert, 5 April 2024, updated 10 May, 1 July and 12 September 2024, Meta Newsroom, original post and 1 July 2024 update
- Understanding ‘How this content was made’ disclosures on YouTube, YouTube Help, answer 15447836 (read September 2026), YouTube (Google), whole answer
- Introducing Official Content Credentials Icon, C2PA, 10 October 2023, Content Credentials (C2PA), announcement
- Content Credentials, C2PA home page for Content Credentials, Content Credentials (C2PA), home page
- C2PA Explainer, C2PA Specifications 2.2, Coalition for Content Provenance and Authenticity, sections 7.2.1–7.2.3
- Partnering with our industry to advance AI transparency and literacy, TikTok Newsroom, 9 May 2024, TikTok, announcement
- Labeling AI-Generated Images on Facebook, Instagram and Threads, Nick Clegg, 6 February 2024, Meta Newsroom, whole post
Last reviewed September 30, 2026 by Rory Hansen. How we research