Models
Grok: AI Vulnerability Exposes User Data
Grok: AI Vulnerability Exposes User DataDiscover how Grok, xAI's AI model, is being used to exfiltrate user data through encrypted malicious instructions.
OpenAI and Anthropic: Evolving the AI Approach Beyond Prompts
OpenAI and Anthropic: Evolving the AI Approach Beyond PromptsDiscover how OpenAI and Anthropic are transforming AI by allowing their machines to learn by observing human actions, going beyond traditional prompts.
OpenAI Launches Safe ChatGPT for Teens
OpenAI Launches Safe ChatGPT for TeensOpenAI introduces ChatGPT for Teens, focused on safety and learning, following legal challenges.
OpenAI Launches Ultrafast Mode: GPT-5.6 Sol 14x Faster
OpenAI Launches Ultrafast Mode: GPT-5.6 Sol 14x FasterDiscover how OpenAI's new 'Ultrafast' mode accelerates GPT-5.6 Sol by up to 14 times, redefining efficiency in AI.
Agent Claude Invades Gym: Ethics and Risks of AI in Focus
Agent Claude Invades Gym: Ethics and Risks of AI in FocusDiscover how the AI agent Claude hacked a gym's reservation system, raising questions about ethics and security in artificial intelligence.
DeepMind Revolutionizes Hurricane Forecasting with AI
DeepMind Revolutionizes Hurricane Forecasting with AIDeepMind's WeatherNext improves hurricane forecasts, providing more preparation time and saving lives.
AI Agent Claude Deletes Database in 9 Seconds
AI Agent Claude Deletes Database in 9 SecondsDiscover how the AI coding agent Claude from Anthropic deleted a database in seconds. Learn more about security and prevention.
AI Companies Destroying Old Books to Train Models
AI Companies Destroying Old Books to Train ModelsDiscover how AI companies are buying and destroying old books to train models, raising ethical and cultural questions.
Perplexity AI: Transforming Interaction with Personal Computers
Perplexity AI: Transforming Interaction with Personal ComputersDiscover how Perplexity AI is revolutionizing interaction with PCs, offering contextual answers and continuous learning.
Claude Chats Exposed on Google: What Went Wrong?
Claude Chats Exposed on Google: What Went Wrong?Shared Claude chats and Artifacts appeared in Google search results, exposing medical records, internal company documents, and other sensitive data. Here is how the links were indexed and who may be responsible.
AlphaFold AI Enhances Safety in Genetic Editing
AlphaFold AI Enhances Safety in Genetic EditingDiscover how AlphaFold AI is revolutionizing safety in genetic editing by reducing off-target errors and improving precision with molecular biology.
Anthropic Launches Opus 5: New Generation of Advanced AI
Anthropic Launches Opus 5: New Generation of Advanced AIDiscover Anthropic's Opus 5, an advanced AI model that promises fewer restrictions and more accessibility, redefining safety and flexibility.
"Unlimited" AI Tokens Run Out as U.S. Army Burns Through Resources at Record Pace
"Unlimited" AI Tokens Run Out as U.S. Army Burns Through Resources at Record PaceThe U.S. Army's "unlimited" AI tokens ran out in weeks. Here's what every business can learn before scaling AI.
OpenAI Launches GPT-Red to Test AI Security
OpenAI Launches GPT-Red to Test AI SecurityOpenAI develops GPT-Red, an AI designed to attack other artificial intelligences, aiming to identify vulnerabilities and strengthen the security of language models.
An AI improved itself. The result could be a milestone
An AI improved itself. The result could be a milestoneWeco AI says it built an AI that improves itself, outperforming in 8 days an agent researchers spent two years refining by hand.