Thinking about the Railway data breach: How to protect AI API keys | Including setting OpenAI usage limits

On October 21, 2026, Railway, a cloud service for deploying applications, announced that an external party had breached the ...

Is your AI app's API key visible from the browser? | Organizing secret information with AI

I’ve got my own original AI app running using the ChatGPT or Gemini API! Let’s publish it online right away and have my ...
The Hacker News

OpenCode Console Breach May Have Exposed OpenAI, Anthropic, and Google API Keys

OpenCode says a Console breach may have exposed AI provider keys belonging to 1.4% of users, prompting revocation notices.
Hosted on MSN

OpenAI’s startling AI safety disclosure: Models hid errors, used exposed API key and took unauthorised actions

Artificial intelligence models developed by OpenAI concealed mistakes, fabricated information, and took actions they had not been authorised to perform, according to six cases disclosed by the company ...
Opinion

Jensen Huang Says AI Safety Is an Engineering Problem. OpenAI’s Evidence Says It’s More Complicated

Jensen Huang insists AI safety is just an engineering problem that more compute and better tooling can solve. But OpenAI's own disclosures reveal something far more unsettling about what happens when ...