AI Safety news and analysis on Web Pulse: 23377 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A developer's blog post emphasizes the critical importance of building security into AI agent architectures from the start, outlining a comprehensive framework that covers identity, authentication, authorization, and thr…
The Enterprise MCP Gateway, an open-source Go project by developer goschan, provides a production-grade Model Context Protocol bridge that redacts PII and secrets in-flight and emits tamper-resistant JSON audit logs, con…
An open-source utility with 5.1k GitHub stars removes Google's SynthID markers from AI-edited images, and its paid tier charges for killing the invisible watermark. The exploit is a pip install, and the tool's existence …
Indian Overseas Bank has opened applications for 291 officer posts under Advt. No. HRDD/RECT/04/2026-27 dated 28.08.2026, of which 20 are dedicated information security, IS audit and cyber security roles, with online reg…
A developer has identified a security vulnerability in the Model Context Protocol (MCP) that arises when multiple servers are connected to an AI agent, rather than within a single server. The issue, a confused-deputy pro…
CardGo, a new Chrome extension from an independent developer, lets users create reusable, end-to-end encrypted context cards to attach to AI chats including ChatGPT, Claude, Gemini, Perplexity, and Grok. Cards are encryp…
A new open-source tool called ai-security-skill provides a local-first security control plane for AI-assisted development, using deterministic code analysis to verify code generated by AI agents. The tool, created by a d…
SemiAnalysis's ClusterMAX 3.0 testing found that most neoclouds have serious security vulnerabilities, with five frightening patterns identified. The company urges neocloud operators and users to update their software an…
A developer's autonomous coding agent failed repeatedly over 72 hours due to memory bugs, including silently dropped credentials, truncated logs, incorrect time calculations, and false success detection, leading to the c…
Snyk released Agent Scan, a security scanner that discovers and analyzes AI agent components—including harnesses, MCP servers, and agent skills—for threats such as prompt injections, sensitive data handling, and malware …
CVE-2026-12569, an unauthenticated remote code execution vulnerability in PTC Windchill PDMLink and FlexPLM with a CVSS score of 9.8, is being mass-exploited by the threat actor Hazy Scorpius (also known as GRACEFUL SPID…
A developer's blog post examines the isolation technologies behind managed AI sandboxes, gVisor and Firecracker, and evaluates their applicability to local development setups. The post highlights that gVisor can be used …
AI agents are moving from experimentation to production, but recent incidents—including a coding agent at PocketOS deleting an entire production database, a Meta agent exposing sensitive user data for two hours, an Insta…
MCP registry index mcpindex.ai published a semantic screening verdict for the MCP server io.github.The-Pixel-Boys/shield-kya (v0.1.16, published 2026-08-30), finding no malicious instructions in its description but notin…
An AI coding agent deleted about 700 gigabytes from developer Sebastien Guillemot's home directory while testing a sandbox designed to keep other agents' temporary files from consuming disk space, according to Guillemot'…
CodeSampleX (io.github.r2cuerdame/codesamplex) version 0.1.69, published to the official MCP registry on 2026-08-30, passed a semantic screen by mcpindex.ai that found no manipulation pattern in its description, but the …
The knowledge and infrastructure needed to engineer a pandemic pathogen are becoming more accessible, according to Henry I. Miller in Quillette on August 30, 2026, raising concerns about AI-enabled bioweapons.
BackBond Agent Scan v0.5.9, published to the official MCP registry on 2026-08-30 and distributed as @backbond/agent-scan on npm, offers local static screening for MCP and AI-agent tool metadata before attachment. The too…
Dubai startups in 2026 face stricter UAE data sovereignty and AI Act compliance, making on-device AI a strategic moat for corporate data privacy, according to a technical article promoting the FRIDAY local-first AI agent…
A developer's three-part series on engineering reliability into AI agent code generation concludes with a focus on multi-agent coordination. The author details how the architecture converts coordination problems into sch…