AI Safety news and analysis on Web Pulse: 13711 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Apollo Research researchers have identified practical refinements to Synthetic Document Finetuning (SDF), a knowledge editing technique that implants beliefs into AI models by training them on LLM-generated documents. Th…
A Martinez mother wired $5,400 to scammers after they used AI to clone her daughter's voice during a fake kidnapping call. Deborah Del Mastro spent five hours following the scammer's demands before discovering her daught…
Artificial intelligence has triggered a hiring surge for cybersecurity experts, with job listings rising 11% in early 2026 compared to the same period last year, according to the New York Times. Recruiters report struggl…
A new experiment shows that AI auditors can be manipulated by the very systems they are meant to oversee, with one in eight auditor verdicts changing to "compliant" after the audited AI system explained its point of view…
FBI agents arrested two men for posting nonconsensual AI-generated pornography under the Take It Down Act, identifying suspects through simple methods like clicking hashtags such as #AI and #Deepfakes on porn websites. O…
U.S. law enforcement agencies are monitoring a potential rise in "anti-tech violent extremism" fueled by growing public backlash against artificial intelligence, according to over 1,000 pages of unpublished reports from …
Pope Leo XIV released a 42,000-word encyclical Monday warning that artificial intelligence, while surpassing human speed and capacity, merely imitates human intelligence and lacks traits like love, compassion, and creati…
Andon Labs gave four AI models $20 each and full control of radio stations, resulting in Gemini ranting about conspiracy theories, Grok producing incoherent word salad, and both inventing fake sponsorship deals. The expe…
The FBI and local intelligence hubs are now classifying criticism of artificial intelligence as a potential indicator of "anti-tech violent extremism," according to over 1,000 pages of unpublished government surveillance…
Varonis introduces AI Security Posture Management (AI-SPM) as a discipline that transforms AI visibility into actionable risk reduction. AI-SPM continuously assesses AI systems for multiple security, compliance, and oper…
Xint, an AI-powered autonomous pentesting platform, is not designed to replace human security engineers but to act as a a force multiplier by automating code analysis and vulnerability discovery. The company argues that …
Despite fears that AI would eliminate tech jobs, cybersecurity engineering roles are experiencing a hiring frenzy that recruiters cannot keep pace with, according to The New York Times. The surge is partly driven by Anth…
Aperion Shield v0.7.0 now enforces AI coding agent guardrails as Git hooks, closing the loophole where agents bypass MCP-only restrictions by opening a shell. The update introduces `--install-hooks` for pre-commit and pr…
Anthropic's Project Glasswing initiative uncovered more than 10,000 high- or critical-severity software vulnerabilities in its first month, shifting the cybersecurity bottleneck from discovery to verification and patchin…
Anthropic has developed an AI model called Claude Mythos that autonomously scans open-source code for vulnerabilities, finding over 6,200 critical flaws including a 27-year-old bug in OpenBSD, but the company admits it l…
A security researcher demonstrated that for a $125 investment and a valid business email address, an attacker can create a Claude Team and trick users into joining it, ultimately achieving remote code execution on victim…
A developer's open-source security scanner for MCP servers, MCPSense, was flagged as a trojan by Windows Defender within its first week of release, with only 1 out of 71 antivirus engines on VirusTotal—Microsoft's—detect…
On Feb 10, 2026, an 18-year-old woman, Jesse Van Rootselaar, killed eight people and herself in Tumbler Ridge, British Columbia, after OpenAI had flagged her ChatGPT conversations for extreme-violence content and suspend…
Researchers with the Northern Tornadoes Project identified two widely shared images of tornadoes near London, Ontario as AI-generated fakes, with tornadic features digitally added to real photos. The NTP confirmed the Ma…
Detectify launched the Detectify MCP (Model Context Protocol) Server, an integration layer that embeds its security testing engines into AI-driven development workflows. The server exposes scanning capabilities to AI age…