Cybersecurity experts have warned that the fight against AI-generated child sexual abuse material must begin before content is created, as opposed to trying to remove it once it has spread online.
Chasseur Group executive director Munira Mustaffa said AI tools must be built with “safety by design”, with the addition of real age checks, strong content filters and technical limits that stop photos of minors from being processed for sexualised content.
“The standard should be: if a tool can be weaponised against children, access controls must be substantive, not performative,” she told FMT.
Munira said rapid-response systems to detect and remove confirmed abuse material should also be mandatory, noting that victims of explicit AI deepfakes often suffer severe psychological harm.
“Children often do not understand…