The Information Machine

🟡 Fear and building

Semafor Technology · Semafor Technology · 2026-07-01

Semafor's technology newsletter reports Anthropic's reconciliation with the US government and reinstatement of its Claude Mythos 5 model export access, Meta's pivot into cloud computing, and Europe's growing risk of AI-driven economic marginalization.

Appears in

Extraction

Topics: anthropicmeta-cloudai-policyeurope-aiai-regulation

Claims

  • The US government lifted its export ban on Anthropic's most advanced models, including Claude Mythos 5, after Anthropic pledged expanded government collaboration and early model access.
  • Meta is building a cloud business to sell excess compute, causing its stock to rise more than 10% as investors welcomed the potential to monetize its massive capital expenditure.
  • Amazon's global greenhouse gas emissions rose from 69.5 million metric tons in 2024 to 80.8 million in 2025 due to data center expansion, and will continue rising for years.
  • Europe has only one near-frontier AI firm and a fraction of US AI compute capacity, putting it at risk of economic decline and eventual irrelevance if it does not build its own AI industry.
  • California Governor Newsom is simultaneously expanding Claude deployment across state agencies and launching a dashboard to track AI-related job losses.

Key quotes

When people like Dario Amodei talk about 'existential danger,' they mean it, just as they mean it when they say that a positive AI outcome could mean an end to climate change, poverty, and disease.
This isn't going to be linear. It's not going to be a one-year story.
Thanks for your partnership on this, Secretary!