AI Safety news and analysis on Web Pulse: 13136 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A writer distinguishes between two strands of AI opposition: one nostalgic for pre-AI systems and reliant on copyright, and a more radical strand that seeks to overhaul the underlying structures enabling AI. The author a…
UK Prime Minister Keir Starmer announced a ban on social media for under-16s, along with restrictions on livestreaming and stranger communication tools, citing child safety. The government also plans to ban AI romantic c…
Russia's security services temporarily disconnected parts of a surveillance system protecting President Vladimir Putin after the US-Israeli killing of Iran's supreme leader, Ayatollah Ali Khamenei, amid fears that AI-pro…
The Trump administration blocked foreign access to Anthropic's Mythos 5 and Fable 5 AI models due to a jailbreak vulnerability, prompting Anthropic to disable the models globally. The move triggered a rally in decentrali…
Tech companies are laying off tens of thousands of workers at a record pace, citing AI as the primary reason, but critics including Marc Andreessen argue AI is a convenient excuse for pandemic-era overhiring. Meanwhile, …
A new article on Towards AI outlines testing and evaluation strategies for production AI agents, focusing on ensuring reliability, accuracy, and trust before failures occur.
Pope Leo XIV's encyclical and President Trump's executive order both address AI safety, but the AI industry's reliance on probabilistic safety is insufficient, according to a new deterministic safety architecture called …
AI bias compounds as models become more intelligent and autonomous, scaling historical inequities from training data, RLHF alignment, and cultural gaps into high-stakes decisions like loan approvals and resume screening,…
Global financial fraud cost victims an estimated $442 billion in 2025, roughly equivalent to Denmark's GDP, according to Interpol's 2026 Global Financial Fraud Threat Assessment. AI deepfakes and fraud-as-a-service kits …
A wealth management assistant's evaluation suite focused on tool routing rather than answer correctness, allowing bugs to reach users. Adding answer-level evals with a three-layer pass criterion caught inconsistencies th…
Forbes published a roundup of five high-profile AI failures at Air Canada, Zillow, Samsung, CNET, and IBM, highlighting governance, data, and trust risks. A Canadian tribunal ordered Air Canada to pay $812.02 after a cha…
Elastic Security Labs released an open-source CI/CD Abuse Detector that uses Anthropic's Claude LLM to analyze pull-request diffs for malicious workflow changes in GitHub Actions, GitLab CI, and Azure DevOps. The tool ru…
A former EMS worker with a traumatic brain injury accuses the entire AI industry of abandoning its safety mission, claiming repeated attempts to present alignment concerns to organizations like LessWrong and Hugging Face…
Cloud AI platforms including xAI, Azure, and AWS are silently swapping models under running workloads, automatically redirecting requests to newer versions without developer notification. The EU AI Act considers such una…
A developer warns that multi-tenant AI systems using vector databases face a specific security risk: cross-tenant data exposure through misconfigured namespaces, shared credentials, or metadata filter bypass. The post de…
A new study reveals that 69% of AI users admit to 'botshitting'—shipping unverified AI-generated work—with heavy users 64% more likely to do so. Researchers warn this behavior leads to moral disengagement, as 40% of work…
The European Commission is assessing the practical implications of a U.S. export control directive that restricts access to Anthropic's latest AI models, Fable 5 and Mythos 5. Commission spokesperson Thomas Regnier said …
Amazon CEO Andy Jassy reportedly raised security concerns about Anthropic's AI models with White House officials, leading the U.S. government to issue an export control directive that forced Anthropic to disable its Fabl…
Three vulnerabilities in LangGraph's checkpointer, disclosed on June 12, allow attackers to chain SQL injection into remote code execution on self-hosted servers using SQLite or Redis backends. The flaws, discovered by C…
Leaders of the Group of Seven (G7) nations are convening in Évian-les-Bains, France, from June 15 to 17 for the 52nd G7 Summit, focused on global economic stability, international security, and the future of AI. French P…