A Google research paper says its four-agent SAFE system is already in early deployment against synthetic spam networks without publishing the performance data behind that claim, Microsoft and YouTube each made their own weekend disclosures, and each one left a gap between what was announced and what was actually confirmed.

Detection Without Disclosure: SAFE Deploys, Posts Policy Tightens

Google surfaced two enforcement changes over the weekend, a new AI spam detector and a stricter Posts policy, both shipped through documentation rather than a full announcement.

Ask YouTube’s Growth Story Gets a Shopping Chapter

YouTube paired a headline usage number with a shopping preview at its Made on YouTube event, though the details behind both remain thin.

Claims Ahead of Confirmation: A Label Test and a Rater Cleanup

One company is trying on new vocabulary before committing to it, while another is cleaning up the human layer meant to keep its models honest.

Quick Hits