OpenAI Seeks $30 Billion in Fresh Funding at $1.25 Trillion Valuation

Morning, {{first name | folks}}! OpenAI is seeking billions in fresh funding at a $1.25 trillion valuation, while Microsoft is testing whether AI can improve complex medical diagnoses. A breach at ConnectWise exposed thousands of organizations to cyberattacks, Meta is facing renewed questions over website content and AI training, and DeepSeek is raising $15 billion to challenge US AI leaders.

OpenAI is seeking up to $30 billion in fresh funding at a potential $1.25 trillion valuation, with UAE investors among those involved in the discussions, according to CNBC. The raise would give the company more capital as it prepares for a possible IPO.

The push reflects how expensive the AI race has become. Beyond building more capable models, OpenAI needs vast computing capacity to serve growing demand, making access to capital and strategic investors increasingly important to its ambitions.

Microsoft’s Diagnostic Orchestrator scored 83% on complex medical cases in a science study, compared with 36% for generalist doctors using reference materials. It coordinates multiple AI agents to investigate symptoms, weigh possible diagnoses, and determine which tests to run.

Instead of relying on one answer, the system works through each case step by step. The results come from simulated cases, however, and real-world hospital testing will be needed to establish whether it can help doctors diagnose patients more accurately.

Hackers linked to China breached ConnectWise’s cloud backup service and used stolen credentials to target around 2,250 Microsoft 365 environments, according to Microsoft. They used AI-powered tools to automate sign-in attempts, reaching nearly 3,000 attempts per minute.

The campaign shows how one breach can give attackers a path to thousands of potential targets. By automating their attempts, the hackers could move far faster than manual attacks, putting pressure on security teams to detect suspicious activity before more accounts are exposed.

Meta has revised its business terms after criticism over wording that appeared to allow it to collect content from websites linked to business accounts for AI training. The company removed the explicit reference to "training" but still allows website content to be used to develop and research its advertising and business products.

That hasn’t put the concerns to rest. Legal experts say the remaining language could still cover AI training, while Meta maintains the update doesn’t change how its web crawlers operate. For publishers, the bigger question is how much control they have over content that Meta collects from their websites.

Chinese AI startup DeepSeek has raised around $15 billion, double its original funding target, with Tencent and CATL among its biggest investors. The company has gained attention for building competitive AI models at lower cost than many US rivals.

The new funding could help DeepSeek expand its computing capacity and develop more models as it competes with OpenAI and Anthropic. It also gives the company more resources to pursue its ambitions as China and the US race to lead AI development.