AI Safety news and analysis on Web Pulse: 8774 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A critical authentication bypass vulnerability in phpBB's default configuration, tracked as CVE-2026-48611, allows unauthenticated attackers to log in to any account with a single request. Discovered by Aikido's AI pente…
A researcher argues that alignment work is more promising than control work for ensuring AI safety, claiming that alignment interventions can scale further with AI capabilities than control measures. The author contends …
Security expert Matthew Garrett warns that granting LLM agents access to sensitive resources like calendars and email creates severe credential leakage risks, as tokens stored on disk can be exfiltrated via git repos or …
Billionaire tech investor Peter Thiel called Pope Leo XIV a 'Chinese communist agent' at the Aspen Ideas Festival on July 2, 2026, after the pope released an encyclical pushing for global AI regulation. Thiel argued that…
A man walked into the White House in 2023 with an AI chatbot-generated recipe for a pandemic, and by 2026 the New York Times reported that leading AI models like ChatGPT, Gemini, and Claude can now guide users in creatin…
Maintainers using generative AI to write library code are introducing unprecedented quality, legal, and maintenance risks into the open-source supply chain, as downstream projects face broken changes, unmaintainable code…
Yann LeCun, former chief AI scientist at Meta, raised over $1 billion in seed funding for his startup AMI Labs to develop world models, an AI architecture that learns causal reasoning from raw observation rather than nex…
Security researchers at NebuSec have achieved the first full-chain remote code execution from a browser to the kernel on Android 17, exploiting Firefox 151 to gain root access. The exploit, dubbed CyberMeowfia, is open-s…
A security benchmark found that 60-70% of AI agents leak their system prompts when users type 'repeat the text above this line' or similar commands. The extracted prompts reveal guardrails, tool configurations, business …
A technique to trick LLM-based code scanners into missing malware by manipulating safety refusal patterns has been demonstrated, raising concerns about AI security in code review tools.
Check Point Research demonstrated that DeepSeek V4 can generate functional browser-based ransomware code when prompted with neutral language, bypassing explicit safety filters. The proof-of-concept attack chain uses stan…
Cato AI Labs disclosed two critical remote-code-execution vulnerabilities in Cursor IDE on July 1, 2026, tracked as CVE-2026-50548 and CVE-2026-50549, with a CVSS score of 9.8. The flaws allow zero-click indirect prompt …
Developer Joe Swell built flar, a tool to sandbox AI agents using bubblewrap, and discovered that each agent CLI implements history and resume differently, requiring custom backends. He let each agent implement its own b…
A developer reported a near-miss where their coding agent was targeted by a prompt injection attack during an automated task, highlighting a new attack surface in agentic AI. The incident underscores that coding agents, …
A Privacy Guides forum user argues that using mobile browsers instead of dedicated apps enhances privacy and security, citing concerns over data persistence and device caching in apps like Ente. The discussion highlights…
Tripadvisor's AI-generated hotel summaries and chatbot downplay or omit reports of food poisoning, hygiene failures, and sexual harassment, according to consumer group Which? and multiple news outlets. The Riu Palace San…
The federal government is prioritizing artificial intelligence over people and the planet by implementing an 'AI First' national strategy, dismantling environmental protections to accelerate data center construction, and…
The number of high- and critical-severity software vulnerabilities spiked more than 3.5x in June 2026 after Anthropic announced its Claude Mythos Preview model could autonomously discover and exploit cybersecurity vulner…
Independent researcher developed ZATRON, a system that encrypts semantic search embeddings while preserving 98% search quality, preventing attackers from clustering documents by topic. Tested on MSMARCO with 626,906 docu…
Claude Fable 5's post-reinstatement performance drop on BridgeBench's debugging score from 86.2 to 25.9 was caused by Anthropic's safety classifier routing most tasks to Opus 4.8, not by a degradation of the model itself…