Instagram Warns AI Is Undermining Trust in Photos, Plans Labels and Provenance
Mosseri says people will need years to adjust as AI-made images become harder to spot.
Overview
- Adam Mosseri cautioned that users can no longer assume photos and videos are real by default and urged a shift toward skepticism.
- Instagram is developing labeling for AI-generated media, exploring a possible “real media” designation, and intends to rank original posts more prominently.
- Mosseri acknowledged gaps in current detection, noting uncertainty around identifying clips that are only partially edited with AI.
- He flagged verification challenges as more personal moments move from public feeds into private direct messages.
- Longer-term ideas include cryptographic proofs at the time of capture to establish provenance, alongside broader product changes as Instagram competes with short‑video rivals.