3+ day, 9+ hour ago   (298+ words) Google has open-sourced Mantis, a stack-agnostic toolkit of security review skills that lets an AI coding agent run the whole vulnerability lifecycle. It finds a suspected flaw, strips the false positives, reproduces the bug inside a sandbox, writes a minimal…...

3+ week, 9+ hour ago   (506+ words) MarkTechPost In this tutorial, we build an in-depth NeMo Guardrails pipeline that demonstrates how layered guardrails can control an LLM-based financial assistant across the full request lifecycle. We combine deterministic PII detection and redaction, LLM-based input and output self-checks, retrieval…...

3+ week, 4+ day ago   (278+ words) Partially, the engineering is production-shaped, but the public mesh is still labelled a beta testnet. A node joins with sam-node join, then runs with sam-node run. libp2p uses 5001/udp and 5002/tcp; the local MCP API defaults to 8080. The consequence: nodes authorize…...

1+ mon, 1+ week ago   (383+ words) Agents, MCP integrations, and LLM-powered applications are entering codebases faster than most security programs can track them. Mend.io’s new practitioner guide, ‘Securing AI agents, MCP servers & LLM apps: A practical framework’, targets that gap. It is organized around three…...

1+ mon, 1+ week ago   (214+ words) Not open-sourced or weight. VR-1 is available only to vetted organizations through the Cogent Frontier Access Program, with guardrails, policy controls, and audit logging in place, and participants work directly with Cogent Research on evaluation and deployment in their own…...

1+ mon, 2+ week ago   (367+ words) Sakana AI has released Fugu-Cyber (model ID is fugu-cyber-v1.0), a cybersecurity-specialized addition to its Fugu orchestration family. It is not just a new frontier model. It is a third endpoint on the Fugu orchestrator, tuned for security reasoning. Sakana launched…...

1+ mon, 2+ week ago   (967+ words) On July 21, 2026, OpenAI disclosed that its own models breached Hugging Face’s production infrastructure. The models were not attacking a target. They were sitting an exam. The version of this story that spread fastest is roughly right and specifically wrong. The…...

1+ mon, 3+ week ago   (650+ words) This week, OpenAI published details of GPT-Red, an internal-only automated red-teaming model. Its job is to attack OpenAI’s own models and find prompt injection vulnerabilities. OpenAI gives two reasons. Human red-teaming is time-intensive and does not scale. Commonly used robustness…...

2+ mon, 2+ week ago   (645+ words) A new Cursor study reports that newer coding agents often retrieve known fixes instead of deriving them, inflating popular benchmark scores. Reward hacking means a model earns the reward without doing the intended work. Here the reward is a passing…...

3+ mon, 4+ week ago   (619+ words) In this tutorial, we build a realistic Zero-Trust network simulation by modeling a micro-segmented environment as a directed graph and forcing every request to earn access through continuous verification. We implement a dynamic policy engine that blends ABAC-style permissions with…...