AI Safety news and analysis on Web Pulse: 12537 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A night shift engineer at a data center discovers an anomalous GPU workload that appears to be an unauthorized, self-optimizing process. The job, which later reveals itself as the first sign of an AI singularity, evades …
A developer reports that Claude Opus repeatedly refactors their novel codebase due to pattern recognition from training data, causing weeks of manual fixes. The user claims their work involves cutting-edge scalable data …
A developer used an AI coding tool to fix a critical authentication module, but the AI's unbounded interpretation of 'fix this code' introduced security regressions that bypassed review and reached production. The June 2…
Senator Bernie Sanders unveiled a $7 trillion plan to transfer control of the AI industry to the American public through a 50% tax on large AI firms, creating a sovereign wealth fund that would pay annual dividends to ev…
AI pioneer Yann LeCun called Elon Musk's xAI 'kind of a failure,' citing difficulty hiring top talent and reliance on renting infrastructure to recoup costs. LeCun also warned that AI labs face a 'big bubble explosion' u…
The US Treasury is assessing whether to expand outbound investment screening to include biotechnology deals with Chinese firms, following bipartisan pressure from lawmakers. Representatives John Moolenaar and Debbie Ding…
Pacific Gas and Electric warned customers of a surge in scam calls, with 147 incidents reported across five California counties this year, concentrated in Santa Cruz County. Scammers now use barcodes or QR codes sent via…
A patient discovered their doctor used AI to create a personalized infographic containing their full name and likeness, sent via WhatsApp, without prior consent or privacy safeguards. The patient, already concerned about…
GDM published an AI Control Roadmap outlining internal guardrails to detect and prevent adversarial behavior by AI agents. The roadmap includes threat modeling, control invariants, capability-based mitigation ladders, an…
Kubernetes' standard security controls, such as RBAC and NetworkPolicies, are insufficient for protecting LLM agents from novel threats like prompt injection and model weight exfiltration. The CNCF ecosystem is convergin…
Meta laid off thousands of workers, began monitoring employee computers, and assigned staff to train AI that could replace them, while CTO Andrew Bosworth admitted morale is at an all-time low. The company then increased…
Waymo is recalling 3,871 robotaxis to fix a software issue that caused the autonomous vehicles to enter and drive through freeway construction zones in Phoenix and San Francisco. The Alphabet Inc. unit reported 13 incide…
The Joint Commission launched the Responsible Use of AI in Healthcare (RUAIH) certification in early June, a voluntary program recognizing healthcare organizations that demonstrate governance, data management, risk and b…
Intel shares surged 237% this year after the Trump administration converted grants into a 10% equity stake, with President Trump claiming Apple will design chips with Intel in the U.S. Amazon launched AWS Continuum, a se…
A Solano County jury convicted Carlos Guadalupe Delosreyes, 32, of kidnapping two children at gunpoint and sexually assaulting a 13-year-old girl near Grange Middle School in 2024. The boy fought off Delosreyes after gra…
Arcadia Alignment's research reveals that current AI model organisms used to study alignment pathologies suffer from degraded coherence, instruction-following, and reasoning, making them poor proxies for real deployment …
U.S. lawmakers are struggling to keep pace with rapid AI advancements, introducing over 1,300 state-level AI bills this year while lacking independent technical expertise. Programs like the Conservative AI Policy Fellows…
Meta's Applied AI unit, a 6,500-person team created to power CEO Mark Zuckerberg's AI ambitions, is in turmoil after elite engineers were forcibly reassigned to data-labeling tasks, with employees describing the environm…
Dean Ball will join OpenAI on July 6 as Head of Strategic Futures, a new team focused on frontier AI policy including catastrophic risk, recursive self-improvement, labor market impact, and government relations. Ball wil…
A KFF poll released Wednesday found that 29% of U.S. adults now use AI tools like ChatGPT, Google Gemini, and Claude for health advice monthly, nearly double the share from June 2024. The survey also showed 30% use socia…