AI Labs Roll Out New Tools for Security, Pictures, and Money

Voters turned out in big numbers and made it clear they want stricter rules for AI and data centers. The midterm results gave Democrats a strong edge, and that shift is shaking up Wall Street.

OpenAI rolled out a new version called GPT-6 Cyber that is built for security work. It helps experts spot weak spots in code and keeps the AI inside a safe box. Only people who are part of the Daybreak program can use it, and groups like Cisco and Mozilla are already testing it.

OpenAI also tweaked its image tool, calling it ChatGPT Images 2.6. The new version runs about half as fast as before, which is great for folks making many picture variations. It can now turn static pictures into short looping GIFs, and the feature works everywhere for all ChatGPT users.

Anthropic teamed up with Plaid to let Claude look at your bank and credit card data. If you link your accounts, the chatbot can break down your spending and suggest ways to hit savings goals. It can’t move money itself, and the company says it never uses your chats to train its models. Right now the feature is a test only for paying members in the United States.

Both OpenAI and Anthropic say they have strong safety steps to protect user data. Still, anyone sharing personal or money details with an AI should think twice. The election buzz shows people want more control over how tech companies handle their information, and the market is watching closely for any sign of AI’s long‑term profit promise.