Urgent.News

What's breaking now, across thousands of outlets.

AI

Microsoft archived PyRIT (Mar 2026) - what LLM red-teamers should use instead

Quick one: if PyRIT (Microsoft's Python Risk Identification Tool) is on your shortlist for LLM red-teaming, check the repo first. Azure/PyRIT was archived on GitHub on March 27, 2026. It's read-only now: no commits, no releases, no issue triage, nothing. Whatever version you pip-installed is the last version you'll ever get. That matters more for PyRIT than it would for most tools, because PyRIT…

Microsoft archived PyRIT (Python Risk Identification Tool) on GitHub on March 27, 2026, making it read-only with no further updates. PyRIT was a framework for custom attack sequences, not a turnkey scanner. If you were using PyRIT for a broad, actively maintained app-layer scanner, you should use promptfoo, which offers 50+ red-team plugins, OWASP/NIST/MITRE ATLAS report mappings, and regular releases.

If you needed model-layer testing, such as jailbreaks, encoding tricks, or data leakage on the base model, garak is a good option. Giskard offers OWASP-mapped detectors and a paid tier for continuous scanning. For a fast, zero-setup smoke test, sentinel-scan-cli, a dependency-free CLI in Python and npm, has 15 attack patterns mapped to OWASP LLM Top 10 categories.

While none of these replace PyRIT's specific multi-turn orchestration, they offer alternatives for general LLM testing.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

OpenAI Faces Safety Scrutiny After Introducing 'ChatGPT For Teens'

This week OpenAI launched a version of ChatGPT exclusively designed for teenagers between the ages of 13 and 17, reports CTV, "saying this version came with stronger protections around content related…

  • OpenAI launches specialized ChatGPT for teens aged 13-17.
  • Platform includes safeguards against self-harm, violent content, and explicit sexual conversations.
  • Critics question effectiveness and trustworthiness of ChatGPT for Teens.

Japan’s Winning Strategy in the Physical AI Era: Transforming Frontline Excellence into World-Leading Industrial Structures

Japan’s advantage in physical AI lies in converting its tacit frontline knowledge into AI‑ready data, redesigning workplaces for AI‑robot integration, and building secure domestic data infrastructure…

  • Japan's AI strategy focuses on integrating hardware, sensors, and operational design.
  • Capturing frontline workers' tacit knowledge is crucial for AI learning.
  • Japan must establish domestic AI infrastructure for secure data management.

More from Sunday 23 August →