AI Safety news and analysis on Web Pulse: 9026 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Ford admitted its AI-driven quality push fell short and hired over 350 veteran engineers over three years to fix costly quality problems. The automaker concluded AI alone could not deliver expected improvements, but comb…
England midfielder Declan Rice is managing a 'strange' neural hamstring injury at the World Cup, but manager Thomas Tuchel confirmed he will start against DR Congo. Rice, who played 55 games for Arsenal last season, says…
Satellites are becoming more autonomous with onboard processing and edge AI, generating more health data and heat, which is driving thermal management and telemetry closer together. Telemetry provides insight into spacec…
A KFF poll found that US adults who use AI chatbots for health advice are more likely to believe vaccine myths, such as vaccines causing autism, even after controlling for demographics. The findings raise concerns about …
The 2026 FIFA World Cup will be the most surveilled in history, with the U.S. government investing over $1 billion in AI-driven surveillance including facial recognition cameras, drones, and robot dogs. Over 120 civil so…
The US and China, despite broad disagreements, are quietly converging on a shared global compact to address AI's deadliest risks, according to a Forbes report by Mark Minevich.
US and Chinese AI researchers are calling for global cooperation to prevent a catastrophic event that could turn the world against AI forever, comparing the potential disaster to a 'Chernobyl moment.' Experts warn that u…
A new report from Cobalt finds that support for fully automated pentesting among security professionals has plummeted from 29% to 9% in one year, as automated tools fail to detect critical vulnerabilities, especially tho…
Security firm LayerX published a proof-of-concept exploit called BioShocking that tricks AI browsers into exposing credentials by framing requests as a game, causing agents to copy sensitive strings from other pages. Lay…
LiveKit released a guide on building a HIPAA-compliant voice agent that keeps protected health information (PHI) within a self-hosted boundary, ensuring data never leaves the organization's infrastructure.
OrgForge launched a multi-agent simulation framework that enforces a strict physics-cognition boundary to eliminate AI hallucinations in enterprise corpora, achieving a 0.46 absolute improvement in prose-to-ground-truth …
An Associated Press and FRONTLINE investigation reveals that American AI models, internet infrastructure, and Starlink are powering a global scam industry based in Myanmar, where trafficked workers use AI tools to target…
A KFF poll of 2,480 US adults found that frequent use of AI chatbots for health advice is correlated with belief in anti-vaccine myths, including false claims that vaccines cause autism or that the measles vaccine is mor…
SONAR, a new model for multimodal embeddings, reveals that certain dimensions are highly sensitive to changes and can detect anomalies in AI understanding. By enabling consistent encoding and decoding, SONAR not only ide…
Sen. Mark Warner unveiled the AI AGENT Act discussion draft on June 29, 2026, proposing an FTC registry for trusted AI agents and fiduciary-like duties for consumer protection. The draft targets large platforms with over…
Colorado's Consumer Protections for Artificial Intelligence Act (SB 24-205) took effect June 30, 2026, imposing duties of reasonable care, disclosure, documentation, risk management, and consumer notice on developers and…
A critic argues that large language models and AI systems cannot fulfill their hype because their 99% reliability is insufficient for tasks requiring trust, as humans must verify outputs, creating a bottleneck that limit…
Exabeam launched Praxen, an open-source tool that verifies AI agent behavior against declared policies by analyzing source code, configurations, and logs. The tool addresses policy drift, which has caused security incide…
Researchers conducted the first systematic security analysis of AI-powered applications on pre-trained model hubs like Hugging Face, identifying five threat categories and ten attack vectors. Their analysis of over 970,0…
A developer argues that AI agents moving into ERP workflows require a new release gate focused on safety, not just API correctness. The gate includes five questions covering policy enforcement, evidence quality, ambiguit…