OpenAI Announces Custom Instructions Feature Available to All ChatGPT Users


Investors question OpenAI's $852B valuation amid fierce competition from Anthropic, whose revenue surged from $9B to $30B annually, driven by strong demand for its coding tools.....
Anthropic co-founder confirmed briefing the Trump administration on its undisclosed AI model Mythos for its cybersecurity capabilities, while maintaining government dialogue despite suing the Pentagon over supply chain risk classification.....
OpenAI confirmed that its products have been affected by a supply chain attack on the Axios library, but no data leaks or system intrusions have been found. The company has updated the security certification for the macOS application and urges users to upgrade to the latest version as soon as possible to prevent risks.
OpenAI expands its Cyber Defense Program, opening the specialized model GPT-5.4-Cyber to global cybersecurity experts, seen as a response to Anthropic's Claude Mythos. The model is optimized for defense scenarios, supporting binary reverse engineering analysis, helping experts analyze malware and enhance network security capabilities.
An internal memo from OpenAI has been leaked, questioning Anthropic's claimed annual revenue of $30 billion as severely inflated. It accuses Anthropic of inflating revenue using the 'gross method,' and states that its product strength is insufficient to support platform-level competition.