Microsoft has updated its Terms of Service to impose restrictions on its AI offerings, which will become effective on September 30. These changes were officially announced on July 30.
The modified terms introduce regulations specifically concerning “AI Services,” which are defined as services involving or powered by Artificial Intelligence (“AI”) systems as described by Microsoft.
The updated section outlines five key rules and restrictions for Microsoft’s AI services:
1. **Reverse Engineering:** Users are prohibited from using AI services to uncover the underlying components of models, algorithms, and systems. Attempting to decipher and manipulate model weights is not allowed.
2. **Data Extraction:** Unless explicitly permitted, extracting data from AI services using methods like web scraping or data extraction is restricted.
3. **Limits on Data Use:** The AI services and data from them cannot be employed to directly or indirectly create, enhance, or train other AI services.
4. **Content Usage:** Microsoft will process and store user inputs and outputs from AI services to monitor and prevent abusive or harmful usage or outputs.
5. **Third-Party Claims:** Users are solely responsible for addressing third-party claims related to their use of AI services, in compliance with relevant laws.
These changes in Microsoft’s Services Agreement come at a time when alterations to Terms of Service involving AI are generating significant attention. Other companies, like Zoom and The New York Times, have also adjusted their terms to address AI-related concerns.
