Microsoft's Seeing AI App Officially Launches on Android, Supporting 18 Languages


Microsoft plans to partner with Anthropic to offer its self-developed Maia 200 chip, aiming to boost competitiveness in the AI chip market. CEO Nadella claims the chip can improve token generation efficiency per dollar by over 30%, but Microsoft still lags behind Amazon and Google in dedicated AI chips.....
Microsoft AI CEO predicts that within 12-18 months, AI will achieve human-level performance in most white-collar professional tasks like accounting, law, and marketing, enabling full automation. This sparks global tech and workplace debate, heralding the era of 'AI office work'.....
Microsoft is gradually revoking Claude Code licenses for thousands of internal developers, a popular AI coding tool from Anthropic, as it shifts focus to its own GitHub Copilot CLI due to product integration and financial considerations. With June 30 marking the end of Microsoft's fiscal year, this license cancellation is seen as a cost-cutting measure, reflecting internal tool competition and resource optimization.....
Microsoft recently launched the MDASH multi-model agent scanning framework, an innovative code security system. Compared to GPT-5.5 and Mythos models, MDASH excels in vulnerability identification. Its unique feature is that it does not rely on a single model, but instead enhances detection capabilities through collaboration among multiple agents, demonstrating cutting-edge applications of artificial intelligence in the field of security.
On May 13, Microsoft launched the MDASH multi-model agent scanning framework, using a multi-agent collaboration strategy that integrates over 100 specialized AI agents based on different large or lightweight models. These agents handle code preparation and vulnerability scanning to enhance code security detection accuracy and efficiency, revolutionizing traditional single-AI model design.....