Amazon Discovered Child Sex Abuse Content in AI Training Data

Amazon found child sexual abuse material in AI training data and removed it before model training. The RSS snippet discloses only that it was removed before training; it does not disclose the source, scale, or timing. The key issue is source opacity: child safety officials say that hinders law enforcement investigations.

Why it matters: Bloomberg reports a rare data-governance incident at a major AI builder: Amazon found CSAM in training data and removed it before training. HKR-H and HKR-R are strong; HKR-K passes on the new operational fact but is limited by missing source, scale, and timing, so this is high-7x