AI Safety news and analysis on Web Pulse: 9280 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Equixly has built a proprietary AI model trained exclusively for offensive security testing, unlike competitors that wrap general-purpose models in prompts. The company argues that owning the model avoids risks like thir…
PrivacyHawk launched PrivacyHawk Enterprise, a solution that identifies and eliminates shadow IT accounts, abandoned SaaS subscriptions, and forgotten third-party services to reduce breach risk. The tool addresses the in…
Meta has restricted its engineers from using Anthropic's Claude Code and OpenAI's Codex AI coding tools to protect proprietary AI training data and internal codebases from exposure. The policy, effective as of June 2026,…
Public backlash against AI is growing, with polls showing three-quarters of Americans want stricter regulation. Billionaire Mark Cuban warns tech moguls that opposition to data centers reflects deeper hatred of AI and we…
Joanna Stern, author of "I AM NOT A ROBOT: My Year of Using AI to Do (Almost) Everything," found that her year-long experiment with AI changed how she parents her two sons, ages 4 and 8. She emphasizes the importance of …
A developer found that the risk of system prompt leakage in AI agents depends more on the underlying model than the prompt itself. In a small measurement with five different models, disclosure rates ranged from near zero…
Guardrails in AI deployments are deterministic or model-based layers that enforce operational constraints at the I/O level, distinct from fine-tuning. They mitigate threats like hallucination, prompt injection, and jailb…
Memetic Forge has published a practical release checklist for AI voice agents, emphasizing the need for narrow completion boundaries, golden-call testing, and multi-layer scoring to ensure safe and useful production depl…
Kevin Indig argues that selling AI as a replacement for humans, a strategy he calls 'substitution positioning,' wins short-term attention but erodes long-term trust with buyers and employees. Data from New York State and…
A drone nearly struck a United Airlines Boeing 737 landing at Newark Liberty International Airport on Friday, coming within 100 feet of the aircraft. The pilot of United Flight 1513 reported the circular, three-foot-wide…
Europe's aerospace and defense sector faces a defining decade as geopolitical instability and advances in autonomous systems drive a push for technological independence. The conflict in Ukraine has exposed vulnerabilitie…
The US military used AI to select thousands of targets during the Iran war, but a missed intelligence note and disconnected databases led to a missile strike on an Iranian school that killed 120 children. Investigators f…
An AI security company faces a lawsuit for publishing hallucinated vulnerability findings, highlighting a crisis of trust in AI-generated security research. The case follows cURL's shutdown of its bug bounty program due …
US Senators Tim Scott and Bill Hagerty introduced legislation to block AI technology transfers to Cuba and Iran, citing national security concerns. The bill aims to prevent potential military applications of AI by these …
Flock Safety's surveillance cameras, including AI-powered license plate readers and audio-detection poles, are rapidly spreading across the US, logging every car and enabling law enforcement to search footage via natural…
A Livemint survey of 550 Indian startups, venture capitalists, and incubators found that data governance has surpassed AI regulation as the top regulatory concern, driven by the Digital Personal Data Protection Act (DPDP…
A developer building an LLM-based email classifier identified a critical design flaw: the model's self-reported confidence score is used as a gate for autonomous actions, but a confident impersonation could bypass it. Th…
California Governor Gavin Newsom signed executive orders in March and May 2026 to expand AI use in state government, including stricter vendor safeguards and labor market studies, while creating a regulatory lane for Ant…
Booz Allen Hamilton, the largest AI services provider to the US federal government, partnered with OpenAI to develop AI solutions for national security and critical infrastructure. The collaboration aims to leverage gene…
The US government has allowed Anthropic to release its Mythos 5 AI model to approximately 100 trusted US companies and federal agencies, reversing previous restrictions over national security concerns. Meanwhile, a Chine…