Twitter shares draft of its new deepfakes policy for public input through Nov. 27, says it won't ban deepfakes but will label them and warn users before sharing
Twitter last month said it was introducing a new policy to help fight deepfakes and other “manipulated media” that involve photos …
Context & Ripple Effects
Twitter is taking an unusually open route to content moderation: rather than announcing a finished rule, it has published a draft deepfakes policy for public input through Nov. 27, signaling that the definition of "manipulated media" is still contested enough that it wants outside comment before enforcing anything.
The draft stakes out a middle position ahead of rivals — no bans, just labels and share warnings — which sets up a direct comparison with Facebook's subsequent outright deepfake ban, a policy critics quickly flagged as riddled with loopholes around satire and simpler deceptions like the distorted Pelosi video.
First-order effects
- Users posting or encountering synthetic or manipulated media on Twitter will see labels and interstitial warnings before sharing, rather than takedowns — enforcement burden shifts to Twitter's labeling pipeline instead of its removal teams.
Second-order effects
- Facebook's harder-line ban now sits against a label-and-warn alternative, forcing the debate onto where each platform draws the line — coverage of Facebook's rule already highlights exclusions for satire and non-AI deceptions that Twitter's draft may or may not close.
- Public comment periods like this one give researchers, civil-society groups, and rival platforms a formal window to shape the final text, raising the cost of writing narrow definitions quietly.
Third-order effects
- If the pattern holds across Twitter, Facebook, and later TikTok's expanded synthetic-media guidelines, platforms will converge not on whether to govern manipulated media but on divergent taxonomies of it — labels versus bans, AI-made versus simply edited — leaving enforcement gaps at exactly the seams between definitions.
- A durable norm of drafting moderation policy in public could become table stakes for major platforms, trading speed of enforcement for legitimacy and pre-empting regulatory impositions.
The trend: Platform governance of synthetic media is standardizing around disclosure-first rules — labels and warnings over bans — with each network's definitional choices becoming the real battleground.