Frontier Daily (Sep 11): 18 items from Anthropic, GitHub (Hugging Face), OpenAI and more + X supplements

Every day Laojin pulls the official channels and compiles the past 48 hours of AI news into a short brief: curation only, no commentary, with the original link and publish time attached to every item.

1. Education Report: How educators use Claude

Anthropic's education report analyzes 74,000 educator conversations to show how faculty use Claude for teaching, research, and building interactive learning tools.

Source: Anthropic News（官网监测） · 2026-09-10 · Original

2. Updates to Consumer Terms and Privacy Policy

Anthropic updated its consumer terms and privacy policy so Claude Free, Pro, and Max users can opt in to having conversations used to improve models, with a five-year retention period for participants.

Source: Anthropic News（官网监测） · 2026-09-10 · Original

3. huggingface/swift-transformers: Swift Package to implement a transformers-like API in Swift

Hugging Face's swift-transformers repository has 1353 GitHub stars and is written mainly in Swift, providing a Swift Package that implements a transformers-like API in Swift.

Source: GitHub (Hugging Face) · 2026-09-10 · Original

4. How a researcher uses Codex and ChatGPT to search for new antimicrobial molecules

César de la Fuente's lab uses Codex and ChatGPT to search living and extinct genomes for antimicrobial candidates to fight drug-resistant infections.

Source: OpenAI Blog · 2026-09-10 · Original

5. Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra

Cognition launched its new SWE-2 model, described as rivaling Fable 5.1 and GPT-Astra, drawing 184 points and 94 comments on Hacker News.

Source: Hacker News · 2026-09-10 · Original

6. Now everyone can put data to work

OpenAI introduced the Data agent in ChatGPT Work, letting users connect company data, uncover insights, and build interactive dashboards with AI using natural language.

Source: OpenAI Blog · 2026-09-10 · Original

7. Tell HN: OpenAI keeps re-enabling the 'allow training' setting

A Hacker News post with 401 points and 157 comments reports that OpenAI keeps re-enabling the 'allow training' setting, urging users to check whether it has been turned back on.

Source: Hacker News · 2026-09-10 · Original

8. deepseek-ai/deepseek-recipe: GitHub 开源项目

DeepSeek's open-source deepseek-recipe repository on GitHub has 198 stars and is written mainly in Rust.

Source: GitHub (DeepSeek) · 2026-09-10 · Original

9. [AINews] not much happened today

AINews recaps AI news for 9/8–9/9/2026, highlighting Anthropic's deeper assessment of real-world cyber incidents involving Claude, in which four incidents occurred during third-party cybersecurity evaluations mistakenly connected to the internet with safeguards disabled, and METR will run an independent investigation for at least eight weeks.

Source: Latent Space · 2026-09-10 · Original

10. Quoting Calif Research

Simon Willison's blog quotes Calif Research announcing a demo of WeWorm, described as the first zero-click worm spreading through WeChat calls on iOS and Android, with the team using AI to find the bug and write the first RCE exploit in about two days and build the worm in one more week.

Source: Simon Willison's Weblog · 2026-09-10 · Original

11. OpenAI Launches ChatGPT for Financial Services

OpenAI releases a tailored ChatGPT for financial services integrating premium datasets from Daloopa, PitchBook and LSEG with GPT-6 Astra reasoning for research, modeling and client materials.

Source: X (Grok x_search, @OpenAI) · within 48h · Original

12. Anthropic Publishes Detailed AI Threat Intelligence Report

Anthropic releases its most detailed threat intelligence report yet detailing sophisticated misuse attempts of Claude for cyberattacks, influence ops, biology and weapons, and how they were stopped.

Source: X (Grok x_search, @AnthropicAI) · within 48h · Original

13. Meta Launches Personal AI Agent Muse

Meta launches Muse, a personal AI agent powered by Muse Spark 1.3 that gets things done across life, with deep dive into its built-in safety, security and privacy architecture.

Source: X (Grok x_search, @AIatMeta) · within 48h · Original

14. Mistral AI Raises €3B in Series D

Mistral AI announces a €3B Series D, the largest equity round ever by a European tech company, to fuel infrastructure, products and continued frontier model development.

Source: X (Grok x_search, @MistralAI) · within 48h · Original

15. DeepSeek Releases V4.1-Flash Model

DeepSeek releases V4.1-Flash, a 552B MoE model with novel asymmetric encoder-decoder architecture, native multimodal, dramatically smaller KV cache, lower pricing and open-source deployment support.

Source: X (Grok x_search, @deepseek_ai) · within 48h · Original

16. Anthropic Releases AI Economic Impact Model for 2030

Anthropic's Economics team shares a new model exploring how AI could affect US jobs, wages and growth by 2030 across modest to extreme scenarios, with an interactive explorer.

Source: X (Grok x_search, @AnthropicAI) · within 48h · Original

17. Google DeepMind Launches AlphaGenome Atlas

Google DeepMind launches AlphaGenome Atlas, over 30x larger than the AlphaFold Database, to help scientists explore genetic variants and disrupted molecular mechanisms.

Source: X (Grok x_search, @GoogleDeepMind) · within 48h · Original

18. OpenAI Shares AI-Powered Defense Factory Cybersecurity Playbook

OpenAI details how it mobilized 250+ people and its latest cyber models to find and fix vulnerabilities, releasing the Defense Factory architecture and playbook for continuous AI-driven vuln discovery and remediation.

Source: X (Grok x_search, @OpenAI) · within 48h · Original

---

_Sources: official blogs of OpenAI, Google DeepMind and Qwen; official RSS of Simon Willison, Latent Space, Import AI and One Useful Thing; Anthropic site monitoring, top AI discussions on Hacker News and open-source updates from key AI orgs on GitHub. X sources are retrieved live via Grok x_search with per-link verification. This is a source digest — no commentary from Laojin._