AI Safety news and analysis on Web Pulse: 13553 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A federal agent testified Thursday that Jonathan Rinderknecht, the man on trial for allegedly starting the deadly Palisades Fire, used ChatGPT to search for the home address of DoorDash CEO Tony Xu with the stated intent…
A developer on a WordPress maintenance automation project redesigned the system to halt all remaining plugin updates after a core update rollback, after discovering that the original "keep going" design caused false-posi…
An open-source project called Guard Skills has been released as a code quality alternative designed to catch systematic failure modes in AI-generated code, such as hallucinated APIs, mock abuse, and premature abstraction…
A developer has identified critical security vulnerabilities in Retrieval-Augmented Generation (RAG) pipelines, warning that connecting LLMs to live data sources introduces risks including data poisoning, indirect prompt…
A developer completed a SwiftUI app development tutorial and tested the app on a real iPhone, learning the full iOS app-building workflow without joining the Apple Developer Program. The developer also organized app idea…
Jaya Baloo, co-founder and CISO of Aisle, argues that risk acceptance is a dangerous fallacy in cybersecurity, urging organizations to address known vulnerabilities rather than panic about AI-driven threats. In a podcast…
Elastic Security has launched an integration with Anthropic's Compliance API to ingest, normalize, and monitor Claude activity within its security operations platform. The integration ingests over 300 event types from Cl…
Apple expanded its Private Cloud Compute (PCC) to run on Google Cloud using NVIDIA Blackwell GPUs, allowing Apple Intelligence features to process heavy AI workloads on third-party hardware while maintaining privacy guar…
A supply chain attack targeting the open-source tool Understand-Anything (57,000+ GitHub stars) hid an obfuscated payload in `homepage/astro.config.mjs` via pull request #206, which executes automatically during every As…
Anthropic's Claude Fable 5 autonomously debugged a CSS bug across multiple browsers in a 15-step investigation, highlighting the rapid maturation of AI coding agents. OpenAI acquired Ona (formerly Gitpod) to enable long-…
The EU AI Act's Article 50 transparency obligations become enforceable on August 2, 2026, with fines up to €15 million or 3% of global annual turnover. The Trust Identity Protocol (TIP), an open, post-quantum standard wi…
Anthropic released Claude Fable 5 with undisclosed distillation-prevention guardrails, prompting a public apology after Simon Willison exposed the hidden safeguards. Separately, a DN42 network administrator went bankrupt…
A new paper from MiniMax, MaxProof, trains a model to write competition-math proofs and achieves high scores on IMO and USAMO problems, but a cross-verification experiment reveals that among proofs the training verifier …
Semgrep's benchmark for AI-powered vulnerability detection shows its Semgrep Multimodal tool finds 3.5x more true positives than a tuned guided prompt at similar precision and lower cost per finding. The system uses stat…
Anthropic CEO Dario Amodei proposed a regulatory framework on June 10 that would grant the US government authority to block AI models deemed to pose "unacceptable risks," including those enabling cybersecurity threats or…
A developer has published a proof-of-concept attack script that tests whether the kickbacks.ai backend can detect multi-window saturation fraud. The script, called "attack-real.mjs," spawns parallel windows that emit eve…
Anthropic CEO Dario Amodei warned in a Bloomberg interview that AI has already automated roughly 90% of many job functions, but the real danger emerges when it masters the remaining 10% of tasks. Amodei described AI cogn…
Loop Harness, an autonomous loop engineering tool for coding agents, is now available, enabling scheduled Claude sessions to perform tasks like code review and PR creation in isolated git worktrees with a second verifica…
A developer built ZATRON, a system that transforms document embeddings into modular barcodes to enable encrypted semantic search without revealing the original vectors. When the developer trained a neural network to reco…
On the latest episode of Clockwise 660, the hosts discussed their expectations for fixes in the 2027 platform updates, their plans for Intel Macs now that support has ended, and whether they will trust the new Siri AI. T…