AI Safety news and analysis on Web Pulse: 11152 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A developer argues that responsibility for AI decisions falls entirely on developers, not the AI itself, citing incidents such as Air Canada's false chatbot response, Waymo mishaps in Texas, and Amazon outages caused by …
Signal President Meredith Whittaker warned that AI chatbots are not friends or conscious beings, cautioning against treating them as sentient interlocutors. She expressed concerns about granting systems like Microsoft Co…
OpenAI released LifeSciBench, a 750-task benchmark to evaluate AI systems on realistic life science research tasks. Its top-performing GPT-Rosalind model achieved only a 36.1% pass rate, failing nearly two-thirds of the …
Security researchers have released a post-trained LLM specifically for penetration testing, which reportedly found thousands of real zero-days. The model, developed by the Argus Red team, treats offensive security as the…
A developer open-sourced NRT-Defense v0.4.0, an adaptive multi-turn defense framework for LLM agent teams that reduces attack success rates to under 1%. The framework addresses vulnerabilities exposed by Lee et al. (2026…
Stanford's 2026 AI Index reveals that 88% of companies now use AI in at least one business function, but fewer than 10% have fully scaled AI in any function, highlighting a gap between adoption and governance. The report…
The European Union has finalized major revisions to its AI Act, postponing high-risk AI rules until 2027-2028 and exempting industrial applications, while keeping the timeline for general-purpose AI models like those fro…
A developer highlights the growing importance of Data Security Posture Management (DSPM) for cloud security, using a healthcare startup scenario to illustrate how sensitive data can spread across multiple ungoverned loca…
Signal President Meredith Whittaker warned that AI chatbots like ChatGPT and Claude are not friends or conscious beings, cautioning against granting them pervasive access to personal data. She criticized Microsoft AI CEO…
Bengaluru police registered an FIR at Sadashivnagar police station after an advocate complained about a defamatory AI-generated video posted on the official JD(S) Facebook page showing DK Shivakumar. The video, which car…
Google DeepMind researchers audited DiffusionGemma, a text diffusion model, and found it is not significantly less transparent than Gemma in terms of variable interpretability, but algorithmic transparency is lower due t…
A developer released NeuroImprint Detector, an open-source framework that audits PEFT adapters in Federated Learning for backdoors that memorize training data. The tool can detect and reconstruct up to 79% of original sa…
As millions of Americans use AI agents to file their 2026 taxes, tests show chatbots miscalculate refunds by over $2,000 on average, raising questions about liability when algorithms fail. Intuit, H&R Block, and startups…
HSIP, a local identity server written in Rust, launches with Ed25519 cryptographic signing and AI agent governance, offering tamper-proof audit trails for individuals, AI agents, and financial institutions. The open-sour…
Prem AI announced multi-GPU confidential inference is now available in production through Fluso, its secure AI workspace for sensitive workloads. The move targets regulated industries needing hardware-backed data protect…
CyberSentinel AI v3.0, an open-source autonomous cybersecurity platform integrating 33 penetration-testing and threat-intel tools with a provider-agnostic AI layer, launched on June 20, 2026. The platform supports multip…
IONA OS, a Rust-based operating system, has integrated an AI directly into its kernel that performs real-time system optimization without hallucinating. The AI cross-checks facts using causal reasoning and confidence dec…
Insight Enterprises (NSIT) has launched Managed Exposure Defense, a new service designed to combat AI-driven cyber threats. The offering aims to help organizations protect against increasingly sophisticated attacks lever…
A French AI safety policy insider argues that the AI Safety Community overemphasizes visible outsider tactics like press and open letters, while underestimating the impact of invisible insider work within ministerial cab…
A developer built an email client that uses an AI trust layer to reduce phishing risk and reorganize emails by relationships and context, aiming to help non-technical users like their parents avoid malicious emails. The …