The Lab · 2026-03-31 08:27:05 · GitHub Issues
A critical security triage reveals a live Hugging Face API token has been publicly exposed in the repository's training data for at least 18 hours. The token, with the prefix `hf_sUYKuMlbFnJkwGkewyHNlNKbD...`, was found embedded within two key data files: `training-data/sft/consolidated_root_sft.jsonl` and `training-da...
The Lab · 2026-04-10 16:53:04 · Habr
Google кардинально изменил положение дел в мире открытого искусственного интеллекта. Выход модели Gemma 4, состоявшийся 2 апреля 2026 года, был встречен не просто как очередное обновление, а как событие, способное перевернуть расстановку сил. Технический директор платформы Hugging Face Жюльен Шомон отреагировал на рели...
The Lab · 2026-04-13 21:22:24 · Decrypt
Chinese AI lab MiniMax released a powerful new AI agent model, M2.7, touted as rivaling Anthropic's Claude Opus on key coding benchmarks—only to quietly update its commercial license terms shortly after publishing the model's weights on Hugging Face. This swift revision of the legal framework governing a major open-sou...
The Lab · 2026-05-11 09:10:30 · Mastodon:mastodon.social:#infosec
Security researchers have uncovered a fraudulent repository on Hugging Face that impersonated OpenAI's Privacy Filter model and distributed infostealer malware to an estimated 244,000 downloads before removal. The fake repo reportedly achieved #1 trending status on the platform, leveraging the credibility of OpenAI's n...