The news desk
Recent developments in artificial intelligence, with the original sources one click away.
Anthropic has expanded its Claude AI integration with Microsoft Excel, offering advanced data analysis, modeling, and debugging capabilities to Pro plan subscribers.
Sources · Zapier / Anthropic expands Claude AI access with Excel integration for Pro users +15
The popular open-source AI assistant, Clawdbot, has rebranded to Moltbot after a trademark request from Anthropic, facing both rapid adoption and serious security concerns.
Sources · AI Research (Gemini)
DeepMind employees are seeking workplace policies to protect them from immigration officials following an alleged federal agent incident at Google's Cambridge campus.
Sources · WIRED AI
Google's AI podcast features the team behind deploying Gemini, one of the world's leading AI coding models, to billions of users.
Sources · Google AI Blog
Three recent arXiv papers investigate using reinforcement learning to enhance large language model reasoning capabilities and confidence calibration.
Sources · arXiv AI (cs.AI)
OpenAI's new Science team releases Prism, a free tool integrating ChatGPT into a text editor designed for writing scientific papers.
Sources · MIT Technology Review
Researchers publish methods to reduce memory consumption and computational costs in large language model fine-tuning processes.
Sources · arXiv AI (cs.AI)
Amazon AWS demonstrates how AppSync Events can serve as foundation for serverless AI gateway architectures with integrated AWS services.
Sources · Amazon AWS AI
European regulators are investigating X for inadequate controls that allowed widespread creation of sexualized deepfakes using its Grok AI chatbot.
Sources · NYT Technology
OpenAI is actively pursuing scientists as users, three years after ChatGPT's launch transformed everyday digital activities.
Sources · MIT Technology Review
New arXiv papers introduce BibAgent for detecting citation errors, a federated learning framework for LLMs, and a declarative layer for agent systems.
Sources · arXiv AI (cs.AI)
Recent arXiv papers examine fundamental limits of large language models, acceleration techniques for diffusion LLMs, and hallucination detection methods.
Sources · arXiv AI (cs.AI)
OpenAI's ChatGPT has begun incorporating content from xAI's conservative-leaning, AI-generated encyclopedia Grokipedia in its responses.
Sources · TechCrunch AI
Researchers release benchmarks for legal LLM applications, adversarial robustness testing, and automated optimization benchmark generation.
Sources · arXiv AI (cs.AI)
Researchers release benchmarks evaluating multimodal AI capabilities in compositional spatial reasoning, embedded development, and surgical contexts.
Sources · arXiv AI (cs.AI)
Researchers released three arXiv papers introducing benchmarks and frameworks for autonomous driving agents, multimodal emotion understanding, and geospatial AI.
Sources · arXiv AI (cs.AI)
Recent arXiv papers evaluate medical AI models, identify ineffective LLM layers, and test clinical decision-making vulnerabilities.
Sources · arXiv AI (cs.AI)
AWS adds IaC support for Bedrock AgentCore, enabling developers to deploy AI agents using CloudFormation, CDK, and Terraform.
Sources · Amazon AWS AI
Tensions over artificial intelligence regulation in America escalated in late 2025, according to MIT Technology Review.
Sources · MIT Technology Review
Claude Code allows users without coding experience to create programs and apps through text prompts, according to NYT Technology reporting.
Sources · NYT Technology