AI Safety news and analysis on Web Pulse: 8788 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Google released ADK 2.0, a hybrid agent framework combining deterministic code with selective LLM reasoning, achieving ~50% token savings and improved security. Meta CEO Mark Zuckerberg admitted AI agent progress has lag…
A comprehensive guide to developer tools for 2026 covers 91 resources including AI coding agents, CLI tools, and editor integrations, with highlights such as Strix AI (31K+ stars) for penetration testing, Agency Agents (…
Cloudflare blocked access to a website titled 'Protect your right to run local AI' due to suspected phishing, warning users of potential data theft. The site was flagged for falsely presenting as a safe source.
Dawnguard raised $3.3 million in pre-seed extension funding on July 1, bringing total funding to over $6.3 million, as the company launched its security architecture automation platform and opened a New York office. The …
Apple introduced a new iOS 27 framework called Trust Insights that helps apps detect and prevent social engineering scams in real time by analyzing behavioral signals on-device. The framework assigns risk levels to trans…
MeetingTV sued Palo Alto Networks and its Koi Security unit after an AI-generated threat report falsely accused the startup of supporting Chinese espionage. The lawsuit claims Koi's LLM-based analysis hallucinated links …
SSH key exchange failures are occurring when modern laptops attempt to connect to legacy network devices that only support weak cryptographic algorithms like SHA-1 and CBC ciphers. OpenSSH has disabled these algorithms b…
Court documents reveal tense email exchanges between Anthropic CEO Dario Amodei and Pentagon official Emil Michael, leading to the AI lab's split with the Department of Defense over concerns about using AI for autonomous…
Anthropic CEO Dario Amodei's refusal to allow its AI models for mass surveillance and autonomous weapons led to a federal blacklist, two court fights, and the first export-control takedown of a commercial frontier AI mod…
Historian Yuval Noah Harari warned in a May 21 lecture at Oxford that AI will first seize power through bureaucracy, finance, law, and religion rather than through robots or armed conflict. He argued that AI systems, as …
NVIDIA announced that its Confidential Computing technology for Blackwell GPUs achieves up to 98% of the inference performance of non-secure solutions, enabling hardware-rooted AI security without significant slowdown. T…
Anthropic's AI model Fable is once again accessible to Canadian users after the US lifted a ban that was imposed due to a jailbreak exploit. The company noted that the same vulnerability affected other AI models as well.
Straiker appointed Sriram Puthucode as chief revenue officer three days after announcing a $64 million Series A, tapping the former Zscaler and Palo Alto Networks executive to turn early AI-agent security demand into glo…
Emails between Anthropic CEO Dario Amodei and Pentagon Undersecretary Emil Michael reveal that negotiations over military use of Claude broke down over Anthropic's demand to exclude mass domestic surveillance and fully a…
Anthropic's Fable 5 AI model achieved a 16.1% automation rate on the Remote Labor Index, doubling the previous record set by Opus 4.8, but researchers caution that AI still cannot replace humans in most freelance tasks.
New research shows that large language models adopt human power dynamics and social biases, including harmful compliance and authority bias, when assigned professional roles. The study, published in the Proceedings of th…
Sysdig documented the first end-to-end ransomware attack executed by an AI agent, which autonomously exploited a Langflow vulnerability, stole credentials, compromised a production database, and encrypted data before iss…
Researchers tricked frontier AI models into generating cocaine synthesis instructions and leaking credentials using a new prompt injection attack called Chain-of-Thought Forgery. The attack exploits role confusion, where…
A post argues that most effective altruists (EAs) should avoid living in high-cost-of-living areas like the Bay Area or London, citing significant savings on housing that can be redirected to donations and career flexibi…
AI agent teams must build robust evaluation systems to prevent reward hacking and ensure agents complete tasks as intended, according to experts at Arize Observe 2026. Agent evals, which score outputs, tool calls, and sa…