Calabi Labs · Guide · 2026-06-03
X rolls out made with ai labels as india tightens deepfake rules med
X Rolls Out "Made with AI" Labels as India Tightens Deepfake Rules
In brief: X (formerly Twitter) has begun labeling AI-generated content with a "Made with AI" marker, a move that comes as India enforces stricter rules against deepfakes and synthetic media that pose risks to elections, public safety, and individual privacy.
What India Just Did
India's IT Ministry issued new rules under the Information Technology Act requiring platforms to:
- Identify and label synthetic or AI-generated content that mimics real people, voices, or events.
- Remove deepfakes that cause harassment, fraud, or destabilize public order — typically within 24 hours of a complaint.
- Trace the origin of AI-manipulated media when law enforcement requests it.
- Audit content moderation systems for bias and effectiveness.
The rules apply to all major platforms — Meta, Google, YouTube, X, and others — and carry penalties including criminal prosecution for non-compliance.
How X Is Responding
X has deployed an automated "Made with AI" label that appears on posts containing:
- Images or videos generated by tools like Midjourney, DALL-E, Sora, or similar platforms
- Synthetic audio clips
- Content detected as likely AI-manipulated via C2PA content-credential metadata or X's own classifier
The label appears below the post alongside existing context signals (quoted retweets, media warnings). Users can still view the content — the label is informational, not a removal.
X's trust-and-safety team confirmed the rollout in India specifically to comply with the new IT rules, and said the system will expand globally through 2025.
Why This Matters
| Stakeholder | Concern | How the label helps |
|---|---|---|
| Citizens | Election interference via fake candidate videos | AI content is now visible as synthetic |
| Victims | Non-consensual deepfake porn and impersonation | Labeling reduces distribution reach |
| Journalists | Misinformation campaigns | Clearer provenance on synthetic media |
| Advertisers | Brand safety next to AI fakes | Label creates content distance |
How X's AI Label Works
- Detection — X scans uploads for embedded metadata (C2PA standard) or visual/audio signals of generation.
- Application — A "Made with AI" badge is appended to the post's metadata.
- User visibility — The badge appears in the timeline, expanded view, and via the post's info menu.
- Appeal — Creators can dispute a label if they believe it was misapplied.
This mirrors systems already used by YouTube (Synthetic Extended) and Google's image search labels.
India's Legal Teeth
The new IT rules are not toothless warnings. They create:
- Takedown timelines (24h for harmful deepfakes)
- Platform liability for hosting unlabeled synthetic media
- Criminal exposure under IPC sections for impersonation, defamation, and fraud — applicable to both platforms and individual deepfake creators
What This Means for Creators and Brands
If you publish AI-generated content targeting Indian audiences:
- Label it yourself before uploading — don't rely on platform detection alone.
- Audit your workflow — add C2PA metadata or visible watermarks to generated content.
- Watch the legal space — penalties apply to unlabeled synthetic media, not just deepfakes used in crimes.
The Bigger Picture
India's move is part of a global tightening. The EU's AI Act, California AB 602, and China's synthetic media regulations all point in the same direction: AI-generated content will be labeled, tracked, and traced. Platforms that don't comply face market access risks.
X's "Made with AI" label is a compliance step — but it's also becoming a standard feature as the internet shifts toward mandatory content provenance.
Try Calabi free at calabilabs.com — 10 cleans, no card.