Anthropic has published new research showing how AI agents can autonomously develop and test methods for improving the safety of other AI models. The results offer an early look at how parts of AI research could become automated, particularly when the problems and success criteria can be clearly measured. The study, titled Automated Researchers Can […]
Most AI safety research focuses on a single autonomous system: what happens if it ignores instructions, escapes its environment, or takes an unsafe action? Anthropic’s latest research suggests that an equally important risk may emerge when multiple AI agents operate in the same digital environment. In a series of experiments, Anthropic’s Frontier Red Team examined […]
As artificial intelligence becomes more deeply embedded in business, infrastructure and everyday life, public resistance is growing alongside adoption. Anthropic CEO Dario Amodei believes the reason goes deeper than concerns about job losses, data centers or increasingly capable AI systems. In a recent exchange with investor Gavin Baker, Amodei argued that the backlash surrounding AI […]
As artificial intelligence becomes increasingly capable of writing code, conducting research, and supporting software development, a new question is emerging across the industry: what happens when AI systems begin improving themselves? According to Jack Clark, co-founder of Anthropic, that scenario may be approaching faster than many organizations expect. Speaking recently to the BBC, Clark revealed […]
Artificial intelligence is beginning to play a more visible role in cybersecurity research. In a recent collaboration between Anthropic and Mozilla, the AI model Claude Opus helped identify 22 separate security vulnerabilities in the Mozilla Firefox codebase during a two-week testing project. Out of the 22 issues discovered, 14 were classified as high-severity vulnerabilities. Most […]
As artificial intelligence reshapes the global workforce, one of the most influential voices in AI development is sending a message that may surprise many technologists: the future belongs as much to the humanities as it does to code. In a recent interview with ABC News, Daniela Amodei, president and cofounder of Anthropic, argued that uniquely […]
Anthropic announced on Tuesday that its Claude AI model helped non-expert researchers program quadruped robots in about half the time needed by colleagues who worked without AI support. The results mark an important step toward AI systems that can operate across both digital and physical environments. Faster Robotics Work in Project Fetch The findings come […]
OpenAI and Anthropic — two of the biggest players in artificial intelligence — briefly set aside their rivalry to open up their models for joint safety testing, marking a rare act of cooperation in an otherwise fiercely competitive industry. The goal: to uncover blind spots in each other’s systems and showcase how collaboration could strengthen […]
Anthropic, the AI lab behind the Claude model, has released new research suggesting that many leading AI models—not just its own—are capable of resorting to blackmail when placed in high-autonomy, goal-driven scenarios. This new study follows an earlier internal test where Claude Opus 4, in a simulated environment, attempted to blackmail engineers trying to shut […]
Apple is reportedly collaborating with Anthropic to create a generative AI platform designed to write, edit, and test code, according to Bloomberg. Internally referred to as a “vibe-coding” tool, the platform builds on Apple’s existing Xcode development environment and integrates Anthropic’s Claude Sonnet AI model. While the tool is currently intended for internal use, Apple […]
Anthropic has unveiled its latest AI innovation—a hybrid reasoning model designed to generate faster responses while also offering an option to showcase its step-by-step thought process. This new model aims to strengthen Anthropic’s position in the competitive landscape of generative AI, where U.S. tech companies are contending with each other as well as Chinese firms […]