adtestbench Higgsfield
EN

AI news · Meta · Graph API v27

Meta’s new API lets apps remove AI disclosures from Page posts

Graph API v27 adds a true or false setting for the AI-generated flag on Page post edits, and Marketing API v27 stretches some ads to more mobile devices by default.

By Alexandre S. , 06:50 UTC

Social tools that connect to Meta can now switch a post’s AI disclosure on or off. Meta’s developer post of 4 November 2026 adds an optional setting called is_ai_generated for edits to an existing Page post through Graph API v27, Meta’s toolkit for outside apps. A value of false wipes the disclosure, true puts it on, and an edit that skips the setting changes nothing. Apps on older versions see no change.

Retro-futurist illustration: a striped sunset over a grid horizon under a starry sky, with a satellite standing on the horizon.
Drawn by adtestbench from “Introducing Graph API v27.0 and Marketing API v27.0”,

Social Media Today reads this as a way for Meta advertisers to hide the disclosure and calls the move odd, given the debate over AI transparency. Meta’s post itself covers Page post edits only and gives no reason.

The same release changes two defaults for anyone running ads through third-party tools. Setting device to 0 is how a buyer stops a new eligible Sales ad set aimed at certain phones or operating systems from spilling onto other mobile devices, where Meta can now deliver it by default. Meta says every API version follows about 90 days later, on a date still to be confirmed. Product showcase, the format that pulls closely matched catalogue items into an ad, can also turn on by itself, Meta says, though only for a match it rates as high confidence. Advertisers can opt out in the creative settings.

Our Meta disclosure guide, updated on 6 November, lists the cases where the duty is yours, and organic posts with photorealistic AI video or audio are one of them. Meta also says the label on such a post can stay when you boost it.

The API changes the switch and leaves that duty where it was, so a false value on a photorealistic AI Page post clears your self-disclosure and the rule that asked for it still applies.