SERIES
Anthropic, Google and Hugging Face presented AI through IPO paperwork, government discounts, controlled cyber access and reusable research formats, showing distribution being organized before trust is settled.
Microsoft, METR, Google and no-thinking researchers all pushed AI toward monitoring, auditing and deployment, making safety look less like settled understanding than operational control.
Microsoft, Google, Meta and the UK government presented bigger AI systems, cloud protection, open weights and pro-innovation rules as public goods, a pattern that favours integrated platforms.
Google put agents into Search and subscriber products while its own research post described unresolved privacy and security problems, making control look like the unfinished layer.