Microsoft has officially announced the AI Access Principles, the company's 11 overarching initiatives, designed to enable individuals and organizations globally to create and apply Microsoft's AI assets and AI in general for the greater good. The principles were officially announced by Microsoft's Vice Chair and President, Brad Smith, at the Mobile World Congress in Barcelona.
The new principles will reportedly provide context for the new initiatives and investments Microsoft has been announcing and introducing throughout Europe for the last two weeks, such as the $5.6 billion in new AI data center investments and the new AI skilling initiatives that will benefit over a million people.
(Photo: Drew Angerer/Getty Images)
The Microsoft logo is illuminated on a wall during a Microsoft launch event to introduce the new Microsoft Surface laptop and Windows 10 S operating system, May 2, 2017 in New York City.
These initiatives and programs demonstrate how the company intends to implement these ideas not only in Europe but also in the US and other countries.
The 11 principles prove to be dominantly about Microsoft's expansion of its AI assets and infrastructure accessibility, whilst also committing to socially-responsible initiatives, such as using environmentally-focused management systems on Microsoft's AI data centers' energy consumption.
Microsoft's AI Access Principles
Headlining the AI Access principles is Microsoft's accessibility expansion for its AI assets to developers. Microsoft describes the first principle as an aim to help train the upcoming generation of OpenAI models and make it easier for users to access and utilize Microsoft's CoPilot applications in nearby data centers. This includes the company's pledge to assist fellow developers, train, and implement open-source and proprietary AI models, regardless of scale.
The second AI access principle, on the other hand, details Microsoft's promise to provide software application developers worldwide with widespread access to AI models and development tools. an endeavor to enable every country to create its own AI economy.
The third and fourth AI access principles build upon the second, wherein it specifies that AI models hosted on Azure, including the Microsoft Azure OpenAI API service will now be openly available to software developers via public APIs and a common public API for network operators.
The fifth and sixth principles center on developers' added capabilities and assurances. Developers can now reportedly decide whether to sell and distribute their AI models, tools, and applications directly to users or through the Azure Marketplace for deployment and use on Microsoft Azure.
The sixth principle is an assurance that Microsoft is not using any proprietary information or data from the development, construction, deployment, or use of developers' AI models in Microsoft Azure to outperform those models.
According to the seventh principle, Microsoft will make it easy for users of Microsoft Azure to export and move their data to another cloud provider, whereas the eighth principle, Microsoft guarantees that all AI models and apps operating in our AI data centers will have their physical and cybersecurity demands met.
Microsoft's Responsible AI Guarantee
As per the ninth principle, Microsoft is committed to implementing a robust Responsible AI Standard that prioritizes human input in AI design decisions and upholds fundamental values such as equity, dependability and security, confidentiality and privacy, accountability, and openness.
The promise to fund programs aimed at distributing AI skilling globally was covered by the tenth principle. The final principle guarantees that Microsoft will employ AI to enhance environmental sustainability requirements and manage their AI data centers in an ecologically conscious manner.
Microsoft released these new principles weeks after the company also announced an accord to use cutting-edge technology to identify and block damaging AI-generated content intended to influence voters. The agreement outlines eight specific obligations meant to counteract election-related false AI-generated content.
