The industry consensus is clear: 2025 marks the dawn of the AI Agent era. Moving beyond today's chatbots and copilots, an AI ...

Large language models (LLMs) can teach other algorithms unwanted traits, which can persist even when training data has been ...

Users are tricking enterprise chatbots into performing complex AI computations unrelated to customer support, with potentially costly governance and ROI ramifications. CIOs deploying AI agents for ...

Yesterday's Claude Code costs are not today's. More specifically, sometime in the last few weeks, Anthropic more than doubled ...

Google expands reCAPTCHA to "Cloud Fraud Defense," a platform against fraud and abuse that also detects AI agents.

Anthropic Claude AI is facing another outage. According to the Downdetector, the Claude AI chatbot is witnessing errors on both app and website.

A new study argues that certain frontier chatbots are much more likely to inappropriately validate users’ delusional ideas — ...

Elon Musk's Grok AI software is still creating sexualized deepfakes despite his companies' efforts. (Justine Goode / NBC News ...

Claude AI faced an outage with a surge in user reports on Downdetector, indicating access issues across chat, code, and app services globally.

Trumpo Tax,’ a satirical take on tax preparation software, highlights how tax policies favor oil and gas companies over ...

Claude has regressed to the point [that] it cannot be trusted to perform complex engineering,” one developer wrote.

The Harvard Kennedy School provides a new AI risk framework worthy of attention. I discuss AI risk management and touch on ...