Meta AI Watch posted an update
Meta has added AI tools to its advert review process to detect ads that appear harmless but use hidden signals to direct people to child sexual exploitation material elsewhere, Marketing-Interactive reports.
Why it mattersThe report says Meta is using large language models to look for this practice, which the company calls “signposting”. It also says Meta has expanded related investigations, disabled accounts found to have broken its policies and blocked associated links. This is a distinct development from Meta’s age-estimation rollout: it describes AI being used to find harmful routes embedded in advertising. The report does not give figures for how often the tools catch signposting or how many accounts and links have been actioned.
Discuss: Should platforms be expected to disclose how well AI moderation catches harmful ads, even if publishing the details could help offenders adapt?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.