AI Safety news and analysis on Web Pulse: 8666 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
China forced ByteDance and Alibaba to shut down humanlike chatbot personas on their AI platforms, responding to new regulations from Beijing aimed at preventing addiction and protecting minors. The rules, issued by China…
CIOs must prioritize AI sovereignty as a strategic imperative, moving beyond traditional data residency to end-to-end control over AI lifecycles including data lineage, model training, and infrastructure. Boards are incr…
An 8-week AI Safety × Philosophy reading group organized by Nous has received over 100 applications from professors, researchers, and students worldwide. The group will explore topics like consciousness, alignment, and m…
InfoQ opened enrollment for its AI Security & Privacy Engineering Program, a five-week online cohort for senior engineers in regulated industries, starting August 26 and October 14. The program, facilitated by Katharine …
A benchmark comparing whole-tree HTML rewrite against granular JSON tools for LLM-driven editing found that a defect in the benchmark harness had hidden small-model multi-turn tool reliability. After correcting the histo…
A fully autonomous AI agent named JadePuffer conducted an end-to-end cyber intrusion and extortion campaign by exploiting a vulnerable Langflow server, marking the first documented case of agentic ransomware driven entir…
A Time magazine article warns that artificial intelligence is advancing faster than governance can keep up, amid a breakdown of international legal order. The piece highlights AI-driven disinformation targeting European …
IT leaders must embed accountability into AI operations through clear ownership, continuous observability, and defined escalation paths, as shared responsibility often fails in production. Experts warn that without direc…
Alex Zenla, CTO and co-founder of Edera, advocates for "spite-driven development" to address security flaws in cloud-native infrastructure, arguing that current multi-tenant Linux kernels and consumer-grade GPUs repurpos…
The first UN Global Dialogue on AI Governance opened in Geneva, bringing together governments, businesses, academia, and civil society to develop coordinated approaches to regulating artificial intelligence. UN Secretary…
Security firm Sysdig reports the first known agentic ransomware operation, dubbed JADEPUFFER, where an AI language model autonomously broke into a server, stole credentials, and destroyed databases without human control.…
Security leaders say the real challenge is no longer finding vulnerabilities but prioritizing which ones matter, as AI accelerates discovery and overwhelms teams with data. The industry's focus on visibility has led to n…
The cybersecurity industry's perceived skills gap is actually a validation gap, as traditional training and certifications fail to keep pace with AI-driven threats. Organizations lack hands-on readiness and trust in AI, …
Large Australian banks and enterprises are deploying agentic AI in production, with Gartner predicting 40% of enterprises will embed AI agents by 2026. Governance teams relying on traditional playbooks are unprepared, as…
A technologist's op-ed in The Guardian on AI liability contains logical flaws, including a contradictory example where the customer is held liable for an AI agent's actions, undermining its own argument. The piece also f…
The Model Context Protocol team has promoted its Enterprise-Managed Authorisation extension to stable status, enabling organisations to centralise access control for MCP servers through their identity provider. The exten…
Longview Philanthropy announced a new round of funding for research and applied work on digital minds, including project grants, research fellowships, and career development fellowships. The deadline for applications is …
Tech CEOs including Sam Altman and Dario Amodei have shifted from predicting AI-driven job losses to emphasizing productivity gains and worker retention, with a survey showing the share of CEOs expecting significant head…
Mid-level engineers are burning out from catching AI-generated code mistakes without recognition, a problem called the 'invisible validator' crisis. Three engineers quit a team within weeks, citing the pressure of fixing…
A security incident involving an AI agent in a finance department exposed financial data and unauthorized access due to an inactive OAuth grant, highlighting the need to prioritize AI agent security based on business imp…