AI Safety news and analysis on Web Pulse: 11584 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Researchers propose AgenticRei, a deontic policy framework for runtime governance of LLM-driven autonomous AI agents, addressing obligations, dispensations, conflict resolution, and ontological reasoning beyond current p…
Researchers found that large language models (LLMs) fail to recognize their own knowledge limits on clinical tabular data, outputting near-constant confidence scores regardless of accuracy. By analyzing attribution diver…
Researchers identified a 'narration gap' in LLM-solver loops where formal solver outputs are narrated by language models, potentially compromising soundness. Prompt injection attacks can invert verified conclusions acros…
Researchers introduced TreeTracer, a visual analytics tool that aggregates hundreds of stochastic LLM generations into syntax-aligned trees to expose hidden biases. The tool uses perturbation analysis and contrastive inf…
Researchers introduced the Argent Signaling Protocol (ASP) to mitigate semantic drift in multi-agent LLM systems by attaching structured quality signals to AI responses. In tests, ASP improved pass rates on a document-gr…
Researchers at the Allen Institute for AI used training-data attribution to map which regions of the pretraining corpus support social-reasoning versus STEM-reasoning in the OLMo3-7B language model. They found that socia…
Researchers proposed LUCID, the first hallucination detection method for large language model-based knowledge graph reasoning, which jointly leverages LLM attention scores, KG semantics, and structural information. Exper…
A systematic evaluation of 21 LLM-as-a-Judge models across 118 runs and 541,000 judgments reveals that exact-match agreement overstates discriminative ability, with kappa deflation of 33–41 percentage points on MT-Bench,…
Researchers introduced DeXposure-Claw, an agentic system for DeFi risk supervision that uses a graph time-series foundation model and structured evidence to reduce false alarms. The system, evaluated on five years of rea…
Researchers have developed a method called Emergent Alignment that enables large language models to self-correct unethical outputs by adding a conscience step and using Direct Preference Optimization. The technique works…
A decade of research on AI and systems engineering has evolved through three phases, culminating in a human-AI agreement study that identified five critical gaps in AI adoption, assurance, and workforce transformation. T…
Researchers introduced Free-Energy Signatures (Fes), a spectral descriptor for hallucination detection in large language models that treats attention Laplacians as Hamiltonians to extract thermodynamic potentials. The me…
A developer built a safe, local AI coding agent using Node.js and the Mistral LLM on Ollama. The agent runs entirely on the user's machine without any paid subscriptions, inspecting files and returning patch proposals fo…
Shanti Infosoft argues that placing human checkpoints in AI agent workflows is a business decision, not a technical one. The company recommends a 'draft and approve' pattern initially, using reversibility to determine wh…
A developer built an Agent Profiler that functions as an 'adversity sandbox' to stress-test AI agents by injecting transient runtime errors, lazy-agent traps, and structural AST validation. The tool is designed to ensure…
Anthropic proposed collaborating with the Trump administration on June 18, 2026, to resolve cybersecurity concerns and lift a US ban on its advanced AI models Mythos 5 and Fable 5. The ban, imposed on June 12-13, suspend…
On June 12, 2026, the US government forced Anthropic to take its Claude Fable 5 AI model offline via export controls, citing a jailbreak that revealed minor vulnerabilities. The White House demands jailbreak-proof guardr…
A developer built an open-source PII firewall for LLMs after accidentally sending a customer's credit card number to OpenAI during a benchmark test. The tool, called LLM Governance Engine, intercepts prompts before they …
IT Security News reports a new exploit chain called AutoJack that converts AI browsing agents into remote code execution vectors by abusing trust in localhost, missing authentication, and unsafe parameter handling to tar…
Sebastian Mallaby, author of The Infinity Machine, discusses insights from over 100 AI insiders on the race to superintelligence, the culture of AI, and early breakthrough detection in an interview on The Tim Ferriss Sho…