Trustworthy AI capabilities released by Microsoft

Microsoft's latest announcement emphasizes the importance of trust in unlocking the potential of AI for organizations and communities worldwide. The company is committed to developing trustworthy AI that is secure, safe, and private, ensuring customers and developers are protected at every level.

New trustworthy AI capabilities: Unlocking human potential starts with trust

Microsoft continues to focus on Trustworthy AI by introducing new product capabilities aimed at enhancing the security, safety, and privacy of AI systems.

Security remains Microsoft's top priority, as highlighted by the expansion of its Secure Future Initiative (SFI). The recently released SFI Progress Report reflects the company's dedication to security through three guiding principles: secure by design, secure by default, and secure operations. In addition to tools like Microsoft Defender and Purview, new capabilities were announced:

• Evaluations in Azure AI Studio to support proactive risk assessments.
• Microsoft 365 Copilot will soon offer transparency into web queries, aiding administrators and users in understanding how web search enhances the Copilot response.
   

Customers like Cummins and EPAM Systems have already implemented Microsoft’s security solutions to strengthen their data protection and governance.

Microsoft's responsible AI principles

Microsoft's Responsible AI principles, established in 2018, guide the company’s approach to building and deploying AI in a way that avoids harmful content, bias, and other risks. Microsoft has developed governance structures, policies, tools, and processes to ensure AI systems are built, tested, and monitored for safety. The company has introduced new tools to help customers:

• A Correction capability in Azure AI Content Safety to address hallucination issues in real-time.
• Embedded Content Safety, enabling customers to deploy safety features in devices with intermittent connectivity.
• Protected Material Detection for Code, now in preview, aids developers in detecting pre-existing content in public source code, fostering safer collaboration.
   

Organizations like Unity and ASOS have already benefited from Microsoft’s AI safety solutions, and the education sector has seen the impact, with partnerships in New York City and South Australia.

Privacy

Privacy is a foundational element of Microsoft's AI strategy. Building on their long-standing privacy principles, Microsoft announced:

• Confidential inferencing in preview, ensuring end-to-end privacy during AI inferencing.
• The general availability of Azure Confidential VMs with NVIDIA H100 Tensor Core GPUs, enabling secure data processing directly on GPUs.
• Upcoming Azure OpenAI Data Zones for the EU and U.S., providing enhanced control over data processing and storage.
   

Companies such as F5 and the Royal Bank of Canada (RBC) are already leveraging Azure Confidential VMs for confidential AI applications.

Trustworthy AI for all

Microsoft reaffirms the need for AI that people can trust, highlighting the potential for AI to reshape business processes, customer engagement, and daily life. With new security, safety, and privacy features, Microsoft is committed to empowering people and organizations worldwide to achieve more through Trustworthy AI. This commitment aligns with Microsoft's mission to expand opportunities, earn trust, protect fundamental rights, and promote sustainability across its operations.