OpenAI released ChatGPT Wallet on November 13, letting the personal AI agent dot buy things and book hotels for users. Cards are added once and each payment is approved with face or fingerprint recognition. For now it is only open to Pro subscribers in the US, so it cannot be used in Korea.
A developer built a macOS menu bar app that shows which agent session is waiting for the user's reply. The menu bar shows a count like 2/5, meaning 2 of 5 running sessions need a response. The app also tracks CPU and memory per session, and it can save and restore sessions.
TDG, an AI cloud company, is holding an event with Microsoft to explore how businesses can adopt AI agents. They plan to share real-world cases of using Copilot and AI agents.
Samsung's Galaxy Z Fold8 now comes with Gemini Intelligence, an agent-style AI that understands what is on screen and can carry out tasks across about 40 apps. A reseller also launched a promotion on Samsung's mall for the phone.
n8n released Nod, a first preview of a personal AI agent that runs as a separate app on Mac and Windows. Users can hand off chores like canceling subscriptions, booking hotels after approval, and collecting recurring bills. Tasks run on each user's own computer, so users must manage availability, permissions, and model connections themselves.
Microsoft said a recent global outage of Microsoft 365 was caused by a technical network problem, not a cyberattack or an AI agent malfunction. The issue came from earlier work to expand network capacity and surfaced after a recent network change. The company has fixed the root cause and restored service in stages, and says it will publish a detailed post-incident report.
NVIDIA held AI Day Seoul over two days at COEX, drawing more than 2,000 attendees. The event focused on physical AI (AI that works in the real world, such as robots) and AI agents (programs that carry out tasks on their own).
AWS has released its own official emulator that lets developers test several AWS services together on a laptop without connecting to the cloud. It was built by AI agents that checked their work against real AWS behavior. Third-party tools like LocalStack have filled this gap, but LocalStack now needs an account and a paid plan for companies.
Anthropic Claude blog (mirror) · Nov 16, 09:00 KST
The Claude Code team says about 90% of its code is written and reviewed by Claude. They rely on verification loops, where agents check their own work with tests, so people can hand off more tasks at once. The article covers end-to-end, property, and differential tests and how the team keeps its checks reliable.
Together AI and Ollama explain how coding agents can run efficiently with open AI models by reusing cached work. Agents resend the whole conversation on every step, so keeping the start of the prompt unchanged and routing requests to the right GPU matters a lot.
Together AI and Ollama explain how coding agents can run efficiently with open AI models by reusing cached work. Agents resend the whole conversation on every step, so keeping the start of the prompt unchanged and routing requests to the right GPU matters a lot.
Together AI found that more developers now have coding agents build with its platform. To see how well those agents actually worked, it began running automated tests on common tasks, watching each step and checking the results.
Together AI found that more developers now have coding agents build with its platform. To see how well those agents actually worked, it began running automated tests on common tasks, watching each step and checking the results.
At the RSA cybersecurity conference, IBM and HashiCorp executives said security for AI agents is the top challenge, especially managing agent identities. Few vendors offered end-to-end solutions. Experts agreed that the time to solve agent security is now.
IBM says its partner network is key to helping businesses adopt AI at scale. The company announced a new Granite model, a next-generation watsonx Code Assistant, and new tools for building enterprise AI apps and agents. Partners such as consulting firms, resellers, and software vendors help turn AI pilots into working products.