Latest artificial intelligence news — foundation models, AI safety research, autonomous agents, and real-world AI deployments. Updated every 30 minutes from top AI publications.
Anthropic has trained Claude to run a post-response self-assessment that judges whether a conversation shows signs of misuse, flagging interactions for offline human review without interrupting users. Anthropic reports t…
The Motion Picture Association Inc. runs its piracy investigations and content-protection cases on a CRM module within monday.com Ltd.'s AI work platform, linking MPA offices in Asia-Pacific, Europe, Latin America, Canad…
An Accenture contractor connected a secure Department of Defense production system holding controlled unclassified information (CUI) used by multiple intelligence agencies to a personal Claude Code account, and the Anthr…
Mistral AI released Mistral Large 4, a 1-trillion-parameter language model with fully open weights and a permissive licence, on October 6, 2026, undercutting proprietary API pricing. The company reports 92.4% on HumanEva…
A developer built NOCTURNE, an open-source, air-gapped environmental telemetry forensics platform that ingests multi-variate nocturnal IoT time-series data and uses a two-stage local AI pipeline to distinguish mechanical…
Developer Ankit Kumar built Ghost in the Machine, an open-source, offline-capable AI experience for the Hacktoberfest Open-Source AI Challenge that runs in the browser on vanilla HTML, CSS and JavaScript. The project pai…
Microsoft Threat Intelligence reported on Thursday that North Korea's state hacking groups have made AI a routine part of their operations, rating the country's overall adoption as "high" and placing three groups — Jade …
A Roboflow tutorial demonstrates using a small object detection model, the RF-DETR Medium checkpoint fabric-kruoy/18 from Roboflow Universe, as a first-pass filter so an AI agent only sends flagged fabric images to a lar…
Alex Chernysh's open-source Bernstein scheduler shipped v3.23.0 on October 24th at 18:29, according to the GitHub release record, which lists 105 changes since v3.22.0. The release notes say the headline fix enforces cri…
Epsilon's 50-engineer Agentic Platform team built intent-driven UI testing as a managed, multi-tenant service on Amazon Nova Act and the Amazon Bedrock AgentCore Browser Tool, layering plain-English test steps on top of …
Mistral AI released the open weights for Mistral Large 4, nicknamed Le Chonk, on October 29th, following an October 6th public preview. The company said the one-trillion-parameter multimodal model has improved at agentic…
MoonPay announced on October 29, 2026 that Whop has joined PayBox Markets, connecting its marketplace of approximately 300,000 businesses to MoonPay's PayBox AI payment product so AI agents can search listings, assemble …
Kiro, an AI integrated development environment that pairs an AI agent with a codebase, can drive a repeatable PHP-to-Java migration through an eight-phase workflow built on reusable skill files, according to a new techni…
A developer built VaultRelink, an open-source tool that scans local Markdown vaults for broken [[wiki links]] and suggests replacements using the quantized ONNX export of sentence-transformers/paraphrase-multilingual-Min…
An independent developer released BandReady, a free, open-source, offline-first IELTS preparation app that runs as an Electron front end around a Python sidecar bound to loopback only, with no account required. BandReady…
Cisco Talos discovered NIK C2, a Go-based Windows command-and-control framework that embeds an OpenAI model directly into its attack chain to analyze harvested data on the fly, dump credentials, stream keystrokes, and de…
Aviator launched Aviator Verify, a product that lets engineering teams write rules called Invariants once and have a separate AI agent check every pull request — including those opened by coding agents — against them, re…
Zooko announced BLAKE3-Servil on 2026-10-26, a new fork of the BLAKE3 secure hash function codebase that keeps the upstream implementation's speed and adds formally proven assembly kernels. BLAKE3 has a thicker security …
India's Computer Emergency Response Team (CERT-In) ran a full-day cybersecurity exercise on 27 October 2026 for the banking, financial services and insurance sector, with more than 300 officials from over 55 financial in…
Westphalia Labs, an independent evaluation organisation founded to test AI in diplomacy and foreign policy, ran eleven AI models ten times each as Secretary-General of a fictional Regional Cooperation Union drafting a jo…