The MCP protocol is gaining popularity among AI tool developers, but the number of associated security threats is also growing.
AI: Events
How to Safely Update AI Services: Canary Releases Across Multiple Clusters
Infrastructure
We explore how companies update AI services without the risk of widespread outages, and why the canary release approach is becoming an industry standard.
Cursor has released an update where AI agents can now independently run and test code directly on a virtual machine.
AI: Events
Liquid AI Releases LFM2-24B, Its Largest Language Model – And It Runs on a Regular Laptop
Products
Liquid AI has introduced LFM2-24B – its largest language model, featuring an unconventional architecture and the ability to run both in the cloud and on local devices.
Anthropic has updated its policy for safe AI development. We break down what the RSP is, why it's needed, and what's new in version 3.0.
Alibaba DAMO Academy has unveiled RynnBrain, an open-source model for robot control capable of interpreting its environment and making real-world decisions.
Liquid AI has introduced LFM2-24B, a 24-billion-parameter language model capable of outperforming larger competitors with significantly lower memory requirements.
Anthropic sheds light on distillation attacks – a method to copy an AI model's behavior without accessing its code – and discusses strategies for defending against such attacks.
Lab
When Mathematical Symmetry is Asymmetric: How Non-Invertibility Solves Two Cosmic Puzzles
Physics & Space • High Energy Physics
Why do neutrinos have mass, and why is dark matter stable? A new mathematical structure offers a unified explanation by breaking the conventional rules of symmetry.