Microsoft Introduces Trustworthy AI Features in Azure to Combat Disinformation and Enhance Security

Sep 24, 2024, 03:24 PM

Microsoft has introduced new features in its Azure AI Content Safety service to address growing concerns about AI-generated disinformation and deepfakes. The new capabilities include a watermarking feature for DALL·E images and a tool named Correction, which aims to revise AI-generated text that is factually incorrect. These updates are part of Microsoft's broader initiative to enhance the security, safety, and privacy of AI systems, setting a new industry standard for Trustworthy AI. Experts have advised caution regarding the effectiveness of these tools. Additionally, the service now offers a hybrid deployment option for on-device scenarios and a Multimodal API to detect harmful and inappropriate content.