Security Experts Warn Apple's CSAM Scanning Sets Dangerous Precedent
Apple plans to scan iCloud Photos for child abuse material using a new NeuralHash system, but cryptography researchers warn the technology could easily be repurposed for mass surveillance.
Apple introduces a new "NeuralHash" system in iOS 15 and iPadOS 15 that scans iPhone photos for known Child Sexual Abuse Material (CSAM) before the images upload to iCloud. When the system detects a match, it attaches a cryptographic safety voucher to the file, and a manual review triggers only after a specific threshold of matching content is reached. Apple insists this technology features an extremely high level of accuracy and includes manual reviews to prevent innocent accounts from being penalized.
Security researchers express serious alarm over this implementation because they fear the underlying technology could easily be adapted for other purposes in the future. Johns Hopkins cryptography researcher Matthew Green points out that while the current focus is on child safety, the system could eventually expand to scan end-to-end encrypted photos or look for other types of imagery, such as anti-government protests. Green also highlights the risk of hash collisions, where a harmless file could accidentally match a CSAM hash and result in a false flag.
Experts worry that Apple's decision breaks a longstanding industry dam and sets a dangerous precedent for user privacy. Green argues that once Apple establishes this scanning capability, governments will inevitably demand that all tech companies adopt similar techniques. Critics compare the technology to surveillance tools used by repressive regimes, warning that the system ultimately undermines the very encryption and privacy standards that users expect from their personal devices.