Instagram cracks down on AI accounts pretending to be human
THE SO WHAT
Meta is formalizing AI identity labels, which means the engagement arbitrage from unlabeled AI personas is closing. Brands and creators running synthetic influencers should assume stricter disclosure norms and design for lower, but more durable, trust-based engagement.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIMeta planned to cut some teams by 60% and replace the work with AI
A plan to cut some teams by 60% and replace them with AI agents—then walk it back when the tech underdelivered—shows how quickly “AI-native org design” can overshoot reality. If your CFO is modeling headcount compression from agents, pair it with explicit capability gates and rollback plans, not just cost curves.
Applied AIThe US military gets its own ChatGPT today
The Pentagon approving an OpenAI product for unclassified work after months of testing moves LLMs from experiment to sanctioned tool inside one of the most risk-sensitive bureaucracies. Defense contractors and govtech vendors should now assume AI interfaces will be expected in workflows touching unclassified analysis, planning, and documentation.
Applied AIAnthropic sued over alleged theft of ‘tens of thousands’ of songs
Multi‑billion dollar claims over training data move copyright risk from theoretical to line‑item. If you’re shipping generative models into music, media, or enterprise content, you need a documented data provenance story this quarter—not after the first demand letter.
Applied AIChatGPT becomes the first AI chatbot designated under the DSA
The EU treating ChatGPT as a Very Large Online Search Engine under the DSA pulls general‑purpose chatbots into the same compliance bucket as search and social. Any AI assistant with scale in Europe should now plan for search‑grade transparency, moderation, and audit obligations—not lightweight SaaS treatment.