01 Anthropic gives Claude Code more autonomy, with guardrails
A persistent agent, stealth "Undercover" mode, and... a virtual assistant named Buddy?
Related coverage also includes [AINews] The Claude Code Source Leak, so this reads as a broader theme rather than a single post.
The main names in the cluster are Anthropic, with the angle leaning on developer-tools.
This cluster spans 2 items across 2 sources, so it was treated as a cross-cutting theme rather than a stray item.
- 2 items across 2 sources
- Main names: Anthropic
- Focus: developer tools
02 Google pushes deeper into AI policy
Google partnered with the Brazilian government on a satellite imagery map to help protect the country’s forests.
The item stands on its own as a signal from Google AI Blog.
The main names in the cluster are Google, with the angle leaning on policy.
This cluster spans 1 items across 1 sources, so it was treated as a cross-cutting theme rather than a stray item.
- 1 items across 1 sources
- Main names: Google
- Focus: policy and regulation
03 Ars Technica AI pushes deeper into AI
Swiss finance minister filed a criminal complaint over Grok's "defamation."
The item stands on its own as a signal from Ars Technica AI.
The main names in the cluster are Ars Technica AI, with the angle leaning on general AI developments.
This cluster spans 1 items across 1 sources, so it was treated as a cross-cutting theme rather than a stray item.
- 1 items across 1 sources
- Main names: Ars Technica AI
- Focus: general
What moved around the edges
Scientists、新しい論点が浮上
The proof-of-concept system produces psilocybin, DMT, and other compounds in leaves of the tobacco plant, potentially easing pressure on wild species and preserving Indigenous traditions.
404 Media AIGradient、新しい論点が浮上
Gradient Labs uses GPT-4.1 and GPT-5.4 mini and nano to power AI agents that automate banking support workflows with low latency and high reliability.
OpenAI Blog研究動向で新しい論点が浮上
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI might let us build “political superintelligence”:…But turning this into a societal upside requires lots of intentional work…As AI systems get more powerful and broaden their real […]
Import AIモデル開発で新しい論点が浮上
2023 study made a lot of assumptions about future "anticipated LLM-powered software."
Ars Technica AI