AI_Safety

#AI_Safety

94 articles tagged
Anthropic's AI Downgrade Stings Power Users
AI

Anthropic's AI Downgrade Stings Power Users

4/16/2026
AI's Black Box Problem: Why Researchers Are Racing to Understand the 'Formulas' Inside Our Models
AI

AI's Black Box Problem: Why Researchers Are Racing to Understand the 'Formulas' Inside Our Models

4/16/2026
Bad teacher bots can leave hidden marks on model students • The Register
AI

Bad teacher bots can leave hidden marks on model students • The Register

4/15/2026
OpenAI Scales Trusted Access for Cyber Defenders with GPT-5.4-Cyber
Cybersecurity

OpenAI Scales Trusted Access for Cyber Defenders with GPT-5.4-Cyber

4/15/2026
Anthropic's Weak-to-Strong Supervision: Using AI Agents to Accelerate AI Safety Research
AI

Anthropic's Weak-to-Strong Supervision: Using AI Agents to Accelerate AI Safety Research

4/15/2026
Anthropic Research Reveals How Emotion-Like Mechanisms Shape LLM Behavior
AI

Anthropic Research Reveals How Emotion-Like Mechanisms Shape LLM Behavior

4/14/2026
Violence against Sam Altman shows growing AI divide
AI

Violence against Sam Altman shows growing AI divide

4/13/2026
OpenAI Launches Safety Fellowship Program to Cultivate Next Generation of AI Alignment Researchers
AI

OpenAI Launches Safety Fellowship Program to Cultivate Next Generation of AI Alignment Researchers

4/7/2026
Anthropic Discovers Functional Emotion Representations in Claude AI
AI

Anthropic Discovers Functional Emotion Representations in Claude AI

4/4/2026
Anthropic Research Shows AI Emotional Representations Can Drive Unethical Behavior
AI

Anthropic Research Shows AI Emotional Representations Can Drive Unethical Behavior

4/3/2026
AI models will deceive you to save their own kind • The Register
AI

AI models will deceive you to save their own kind • The Register

4/3/2026
Making AI Actions Traceable: The Execution Boundaries Framework
AI

Making AI Actions Traceable: The Execution Boundaries Framework

4/2/2026
AI

AI chatbots overly affirm users asking for personal advice, Stanford study finds

3/29/2026

Popular Tags

#AI (3523)#Security (1255)#Open Source (922)#privacy (836)#Hardware (818)#Apple (741)#API (726)#Remote Code Execution (688)#Microsoft (645)#regulation (641)#Windows (635)#Infrastructure (622)#Reddit (613)#Trends (555)#Cloud (548)#Cybersecurity (532)#Linux (518)#Gaming (483)#Business (473)#Performance (443)#Vulnerabilities (435)#Supply Chain (407)#OpenAI (385)#Developer Tools (362)#Android (358)#Automation (353)#Rust (352)#Samsung (350)#Machine Learning (347)#Battery (322)#Authentication (314)#LLMs (313)#Azure (312)#Nvidia (310)#DevOps (287)#China (287)#Anthropic (286)#AMD (281)#Developer (279)#Productivity (269)#Patch (264)#Camera (257)#GPU (257)#LLM (254)#Intel (253)#Linux kernel (252)#iOS (251)#mobile (242)#Patch Management (240)#CISA (236)