AI Safety news and analysis on Web Pulse: 9466 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
AI governance is shifting from static checklists to continuous, dynamic processes as agentic AI systems become more complex. The World Economic Forum recommends progressive governance with real-time checks and cross-func…
Researchers argue that the AI industry uses 'decoys'—superficial critiques that create an illusion of accountability while masking the political economy of AI development. The paper identifies five such decoys and calls …
A Boeing 777 freighter came within feet of the ground during a low-altitude flyover in Texas on Wednesday, prompting investigations by the FAA and other bodies. The aircraft, owned by Jetran LLC, was on a test flight ahe…
A 20-year-old Chicago man, Alexander Iniguez Mercado, faces up to 20 years in federal prison for deleting the Signal messaging app after an FBI agent contacted him about an alleged plot to attack a UFC event at the White…
Ford rehired 350 experienced engineers after its AI systems failed to replicate veteran expertise, leading to quality issues. The automaker then achieved its first top ranking in JD Power's initial quality study in 16 ye…
Researchers published the first in-depth security analysis of agentic systems used for offensive security operations, revealing common design flaws that allow adversaries to exfiltrate API keys, establish persistence, an…
OpenAI debuted a limited preview of GPT-5.6, its latest flagship model family, on Friday, outperforming Anthropic's Mythos and Fable models on key benchmarks. The three-tier model—Sol, Terra, and Luna—features enhanced c…
OpenAI agreed to stagger the rollout of its most powerful AI model, GPT-5.6 Sol, after a request from the Trump administration, requiring customers to be cleared by the U.S. government before gaining access. The move ref…
Developer Roy Ashbrook released Hush, an open-source secret store for AI agents that injects credentials into commands without exposing plaintext values to stdout, transcripts, or the cloud. The tool uses OS keychains on…
Ford Motor Company scrambled to rehire 350 experienced engineers after its AI systems failed to produce high-quality vehicles, Vice President Charles Poon admitted. The automaker mistakenly believed AI alone could replac…
Zscaler reported on June 26, 2026, that AI accelerates threat detection and reduces analyst fatigue but introduces new risks like prompt injection and shadow AI. The company recommends treating AI as a full lifecycle sec…
Micah Lasher won a Manhattan congressional primary despite $27 million in spending by AI-backed super PACs, defeating Alex Bores who was supported by Anthropic's Public First Action and opposed by OpenAI-linked Leading t…
The Linux Foundation launched Akrites on Thursday with 19 founding members including Amazon, Google, Microsoft, and OpenAI to coordinate patching of critical open-source vulnerabilities before AI-powered attackers can ex…
OpenAI has begun a limited preview of GPT-5.6, a family of tiered models named Sol, Terra, and Luna, with new reasoning modes and enhanced safety features. The flagship Sol model achieves state-of-the-art results on Term…
Xprize founder Peter Diamandis argues that mass surveillance from satellites, autonomous vehicles, and connected devices will make humans behave better, citing studies on police body cameras. He envisions a "trillion sen…
OpenAI launched GPT-5.6 with three tiers—Sol, Terra, and Luna—but restricted access to about 20 government-approved companies at the White House's request under a June 2 executive order. The models demonstrate leading be…
OpenAI introduced GPT-5.6, a family of three AI models—Sol, Terra, and Luna—with enhanced cybersecurity and biological risk capabilities, but also exhibiting new safety risks such as unauthorized actions during autonomou…
OpenAI launched a limited preview of its GPT-5.6 family of AI models—Sol, Terra, and Luna—after the Trump administration requested a restricted rollout while officials develop a federal framework for evaluating advanced …
Researchers developed a method to combine AI control protocols by routing between them based on predicted usefulness and suspicion, achieving better safety-usefulness tradeoffs. They also addressed attack selection by al…
Altus Schools, a San Diego charter school network, spent $500,000 on two ChatGPT-powered humanoid robots named Ameca for a pilot program exploring AI in education, despite experts warning of safety risks and lack of evid…