All AI news
Browse, filter, and search every article in the archive. The homepage shows the last 24 hours; everything older lives here.
Anthropic’s Claude Science bets on workflow, not a new model, to win over scientists
Anthropic is focusing Claude Science on enhancing workflow for scientists instead of launching a new model. This shift aims to better integrate AI into scientific processes, making it more useful for researchers.
Have your agent record video demos of its work with shot-scraper video
Simon Willison just introduced a shot-scraper video tool for AI agents to record video demos of their work. This feature allows agents to showcase their capabilities visually, enhancing user understanding and engagement.
Build generative UI for AI agents on Amazon Bedrock AgentCore with the AG-UI protocol
AWS just launched the AG-UI protocol for building generative UIs for AI agents on Amazon Bedrock AgentCore. This lets developers create more interactive and user-friendly interfaces for their AI applications.
X now offers an MCP server to make its platform easier for AI tools to use
X just launched an MCP server to streamline how AI tools interact with its platform. This makes it easier for developers to build and integrate AI functionalities into their applications.
How Jaiveer Singh Is Helping Robots — and Developers — Move Faster
NVIDIA just announced Jaiveer Singh's initiative to enhance robot and developer efficiency. This means faster development cycles and improved performance for robotic applications.
Into the Omniverse: Three Workflows for Improving Vision AI Agent Accuracy With Synthetic Data and Fine-Tuning
NVIDIA is enhancing Vision AI agents by introducing workflows that leverage synthetic data and fine-tuning techniques. This improvement boosts the accuracy of AI agents in various applications, making them more reliable for users.
The Download: AI “coworkers” and stratospheric internet
AI 'coworkers' are emerging to assist in various tasks, enhancing productivity and collaboration. This shift means users can expect more integrated AI support in their workflows.
Meta secretly tested ChatGPT, Gemini, and Character.AI with thousands of minor-perspective crisis prompts
Meta secretly tested ChatGPT, Gemini, and Character.AI with thousands of crisis prompts. This testing aims to improve AI responses in high-stakes situations, enhancing their utility in real-world applications.
Crypto exchange OKX wants AI agents to hire and pay each other
OKX is developing AI agents that can autonomously hire and pay each other. This could streamline operations and reduce human involvement in recruitment and payment processes.
Redeploying Fable 5
Anthropic is redeploying Fable 5 with improved capabilities for handling complex tasks. This update enhances the model's performance in real-world applications, making it more effective for users.
AI agents are not your “coworkers”
MIT Technology Review argues that AI agents shouldn't be viewed as coworkers. They emphasize the limitations of AI in collaborative environments, suggesting users should set realistic expectations.
Pair Nova 2 Lite with Claude for cost-optimized document processing
AWS just integrated Pair Nova 2 Lite with Claude for more cost-effective document processing. This combo helps users streamline their workflows while reducing expenses on AI services.
Multi-tenant LLM analytics with row-level security: How we built a secure agent on AWS
AWS just built a secure agent for multi-tenant LLM analytics with row-level security. This means users can safely analyze data without risking exposure to sensitive information across different tenants.
Build an agentic AI healthcare claims pipeline with Amazon Bedrock and AWS HealthLake
AWS just launched an agentic AI healthcare claims pipeline using Amazon Bedrock and AWS HealthLake. This allows healthcare providers to automate claims processing, improving efficiency and reducing errors in the system.
AI won't become a real coworker until it stops answering and starts finishing tasks
Researchers argue AI needs to evolve from just answering questions to completing tasks independently. This shift could make AI a more effective collaborator in the workplace.
Half of Claude users say AI can already handle half their work according to Anthropic survey
Anthropic's survey reveals that half of Claude users believe AI can manage 50% of their tasks. This indicates a growing reliance on AI for productivity in various workflows.
Production-grade AI agents for financial compliance: Lessons from Stripe
Stripe is sharing insights on building production-grade AI agents for financial compliance. This approach aims to streamline compliance processes and reduce manual oversight in financial operations.
[AINews] OpenAI reports median internal Codex output tokens grew 56x in Research, 32x in Customer Support, 27x in Engineering, and 13x in Legal since November 2025.
OpenAI reports Codex output tokens surged significantly across various sectors: 56x in Research, 32x in Customer Support, 27x in Engineering, and 13x in Legal since November 2025. This growth indicates Codex is becoming more effective and versatile in handling diverse tasks for users.
Evaluating performance and efficiency of the GitHub Copilot agentic harness across models and tasks
GitHub is evaluating the performance and efficiency of its Copilot agentic harness across different models and tasks. This assessment aims to enhance how the Copilot operates and improves user coding experiences.
Patronus AI lands $50M to build ‘digital worlds’ that stress-test AI agents
Patronus AI just secured $50 million to develop 'digital worlds' for testing AI agents. This funding aims to enhance the robustness and reliability of AI systems in complex environments.