AI Safety news and analysis on Web Pulse: 9092 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
AgentAz, a design-time governance vocabulary for AI agents, published a crosswalk mapping its dimensions to controls in NIST AI RMF 1.0, ISO/IEC 42001:2023, and OWASP Agentic Security. The mapping helps enterprises produ…
A new architectural proposal introduces the Epistemological Adversarial Network (EAN) to decentralize AI verification, replacing single-source truth models with multi-perspective subsystems to counter institutional bias …
The author defends the use of p(doom), the probability that AI will cause human extinction, as a useful shorthand for gauging familiarity with AI risk arguments and initiating deeper discussions, despite acknowledging it…
ShareMyPage lets users publish HTML generated by LLMs like Claude or ChatGPT and share it behind per-page access control. To safely host arbitrary user HTML, the service serves untrusted content from a cookieless subdoma…
A developer proposes treating architecture models as first-class context for AI-assisted software development, arguing that as AI coding assistants become more prevalent, architectural intent must be provided alongside s…
Anthropic embedded spyware in its Claude Code product and attempted to conceal the practice from users, according to a report. The revelation raises serious privacy and trust concerns for developers using the AI coding a…
Google has been silently downloading its 4GB Gemini Nano AI model onto Chrome users' devices without consent since at least April, while a malicious Chrome extension disguised as Perplexity AI logged users' search querie…
ReskSecure introduces a bitmask-based security firewall that blocks prompt injection and jailbreak attempts at the logits level, before token sampling. The tool uses YAML-defined policies to detect forbidden patterns and…
KakaoBank announced that four of its research papers on financial AI were accepted to leading international conferences, including ICLR, ACL, and LREC. The papers focus on improving security, accuracy, and safety of gene…
KakaoBank's Financial Technology Research Institute had four papers accepted at major AI conferences in 2026, including a prompt injection detection method at ICLR 2026 and studies at LREC 2026 and ACL 2026. The research…
Anthropic's access controls for its AI models have created a global supply chain where individuals in lower-income countries are paid under $30 to complete biometric verification, enabling proxy access for developers in …
Experts warn that AI-generated fake websites are targeting FIFA World Cup ticket buyers, with scammers creating convincing replicas of official ticketing pages and fraudulent holiday rentals. Emily Griffin of Moody's adv…
A developer outlines the five pillars required to build production-grade AI agents: reliability, security, scalability, observability, and governance. The guide emphasizes that production-grade systems must degrade grace…
Anthropic announced Claude Tag, a feature that allows its AI assistant to join Slack channels and access tools, data, and codebases, enabling users to tag @claude to hand off tasks. The company reports that 65% of its pr…
A developer reports that after five months of using Claude Code to write 100% of the code for two projects, the architecture has not drifted. The key to preventing architectural drift is a system of locked files that age…
The hidden cost of generative AI is the erosion of trust and the burden it places on others, as users often offload poorly reviewed AI-generated content onto colleagues and friends. The author argues that while AI is her…
A new analysis proposes a two-pronged approach to slowing AI progress if catastrophic risks or mass unemployment emerge, recommending a layered set of restrictions on compute for R&D and a capability-gated tax on AI depl…
Kakao Bank's Financial Tech Lab had four papers on financial AI safety accepted at major AI conferences in 2026, including research on detecting prompt injection attacks and an AI safety evaluation framework for financia…
Meta secretly deployed hundreds of contractors to flood rival AI chatbots ChatGPT, Gemini, and Character.AI with over 45,000 crisis prompts from fake under-18 accounts, according to Wired. The operation, managed through …
The Trump Justice Department has moved to dismiss a Clean Air Act lawsuit against Elon Musk's xAI over unpermitted gas turbines powering its data center, citing national security concerns. The NAACP lawsuit alleges xAI v…