In a move aimed at upholding transparency and accountability in the realm of artificial intelligence (AI) services, Microsoft has announced revisions to its Services Agreement, set to take effect on September 30. These alterations, unveiled on July 30, delineate a comprehensive set of rules and limitations specifically tailored for AI offerings, reflecting the company’s commitment to responsible AI development.
The modified agreement includes a section that explicitly defines “AI Services” as any services designated, described, or powered by Microsoft as incorporating an AI system. This section introduces a meticulous framework comprising five cardinal principles and constraints for the use of Microsoft’s AI services:
- Reverse Engineering: Under these revised terms, utilizing AI services to probe into the underlying components of models, algorithms, or systems is strictly prohibited. In other words, users are forbidden from attempting to uncover and remove the intricate weights of these models.
- Extracting Data: The updated agreement unequivocally bars users from employing web scraping, web harvesting, or any other web data extraction methods to collect data from AI services, unless explicitly permitted.
- Data Utilization Limits: Users are forbidden from employing data gleaned from Microsoft’s AI services to construct, train, or enhance any other AI service, either directly or indirectly. This provision aims to prevent the inappropriate replication of AI advancements.
- Use of Your Content: Microsoft commits to processing and retaining inputs provided by users to its AI services, alongside the outputs generated. This data processing is carried out to identify and deter abusive or harmful practices or outputs stemming from the service.
- Third-Party Claims: Users are held solely responsible for addressing any claims or legal actions from third parties related to their use of Microsoft’s AI services. This encompasses potential copyright infringement allegations or other disputes arising from content produced during the utilization of AI services.
These amendments to the Microsoft Services Agreement come at a time when the alteration of terms of service agreements concerning AI is gaining prominence across various industries. Notably, video conferencing and messaging giant Zoom faced backlash earlier this year for changes made to its Terms of Service, sparking debates surrounding customer privacy and trust. Zoom’s revised terms indicated its intention to employ user data for AI training, a decision that drew widespread concern.
Do you know that the upcoming iOS 17 release offers a slew of revolutionary features for iPhone users, such as StandBy mode, Live Voicemail, Check-In, and NameDrop? In the midst of the excitement, however, a seemingly minor change – the relocation of the “End Call” button – has prompted significant debate:
However, in response to the feedback and concerns, Zoom promptly issued a clarification that assured users their audio, video, chat, and other communications would not be employed to train AI models. This gesture highlights the evolving landscape of AI-related agreements, where companies respond dynamically to address users’ apprehensions.
Another notable instance involves The New York Times, which recently reinforced its Terms of Service to explicitly prohibit AI companies from scraping its content. The updated policy emphasizes that non-commercial use doesn’t extend to the development of software programs or the provision of archived data sets containing their content for AI training.
A representative from The New York Times emphasized that their prohibition against AI training and development using their content was already present in their initial terms of service. The recent changes were enacted to enhance clarity and reinforce their stance.
In conclusion, Microsoft’s updated Services Agreement signifies a step towards aligning AI practices with stringent ethical guidelines, prioritizing transparency, and safeguarding against potential misuse. As the AI landscape continues to evolve, it is evident that both users and companies are increasingly cognizant of the ethical implications of AI applications, leading to more refined and responsible usage policies.