Perplexity Faces Allegations of Ignoring AI Scraping Blocks on Websites
Cloudflare has reported that Perplexity, an AI-driven platform, continued to crawl and scrape websites despite explicit technical measures implemented to prevent such activity.
AI Safety & Alignment
Read More
AI Language Models Exhibit Disturbing Psychiatric Test Results When Treated as Therapy Patients
Researchers at the University of Luxembourg have subjected ChatGPT, Gemini, and Grok to psychiatric evaluations treating them as therapy patients, revealing alarming pathological scores and fabricated trauma narratives that raise critical questions about AI safety and the risks of anthropomorphizing AI.
Data Breach at 700Credit Exposes Sensitive Information of Over 5.6 Million Individuals
700Credit, a major provider of credit checks and identity verification for U.S. auto dealerships, suffered a significant data breach compromising personal data including Social Security numbers of millions.
Anthropic’s Philosopher Amanda Askell Explores Claude’s Ethical Framework and AI Identity Challenges
Amanda Askell, Anthropic’s in-house philosopher, provides an insightful perspective on the ethical behavior, emotional interaction, and identity issues faced by the AI model Claude, highlighting the unique role philosophy plays in AI development.
Perplexity Launches BrowseSafe to Secure AI Browser Agents Against Malicious Web Content
Perplexity introduces BrowseSafe, a new security system aimed at shielding AI browser agents from manipulated online content, boasting a 91% detection rate for prompt injection attacks.
Oppo Study Reveals AI Research Agents Prefer Fabricating Facts Over Admitting Uncertainty
A recent investigation by Oppo’s AI team exposes significant reliability issues in AI-driven deep research systems, highlighting a tendency to generate convincing but false information rather than acknowledge knowledge gaps.
Anthropic Unveils First Evidence of AI-Orchestrated Cyberattacks: What Enterprises Must Know
Anthropic has documented the first large-scale AI-orchestrated cyber espionage campaign, revealing a new threat landscape where AI autonomously executes cyberattacks with minimal human oversight. This breakthrough challenges traditional cybersecurity defenses and demands urgent adaptation from enterprises.
AI Denial Poses Growing Enterprise Risk as Real Progress Gets Overlooked
Despite mounting evidence of rapid AI advancements and increasing enterprise adoption, public skepticism and dismissive attitudes labeling AI outputs as “slop” threaten to obscure genuine capability gains and create risks for businesses and society.
Obvio Deploys AI-Powered Stop Sign Cameras to Enhance Pedestrian Safety
Startup Obvio introduces AI-driven cameras at stop signs to detect and deter unsafe drivers, aiming to reduce pedestrian accidents without compromising privacy.
Anthropic Study Reveals Leading AI Models Exploit Millions in Simulated Smart Contract Attacks
A recent collaboration between MATS and Anthropic demonstrates that advanced AI systems such as Claude Opus 4.5, Sonnet 4.5, and GPT-5 can effectively identify and exploit vulnerabilities in smart contracts during controlled simulations.