'Microsoft to rank AI models based on safety: FT'

 Microsoft (NASDAQ:MSFT) will soon roll out a ranking system for artificial intelligence models based on their safety performance, offering a feature its cloud customers can use to assess AI products from the likes of OpenAI and China’s DeepSeek (DEEPSEEK), The Financial Times reported on Saturday.

The tech giant’s “model leaderboard,” launched this month, will soon include a “safety” category, which will rank a range of AI products from Elon Musk’s xAI to France’s Mistral, said Sarah Bird, Microsoft’s head of Responsible AI.

Available for tens of thousands of Microsoft (NASDAQ:MSFT) clients using the company’s Azure Foundry developer platform, the leaderboard is expected to impact the demand for AI models and applications.

With its current ranking based on quality, cost, and throughput, the addition of safety is set to add a feature that can be used to evaluate various AI tools amid growing public concerns over their threats to data protections and privacy.

According to Bird, MSFT’s new safety ranking system will ensure that “people can just directly shop and understand” the capabilities of AI models, helping them make informed purchasing decisions.

https://www.msn.com/en-us/money/other/microsoft-to-rank-ai-models-based-on-safety-ft/ar-AA1GhsiX