The agreement prohibits users from deploying Microsoft AI services for nine specific high-risk use categories, including consequential decision-making without human oversight, social scoring, biometric-based categorization to infer protected characteristics, facial recognition database expansion through untargeted scraping, emotional state inference from physical characteristics, and real-time law enforcement facial recognition on mobile cameras.
This analysis describes what Microsoft Copilot's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
These nine prohibited use categories in Section 14.s.ix closely track prohibited AI practices under the EU AI Act and represent contractual restrictions that apply to all users of covered AI services globally. Organizations deploying Microsoft Copilot or other AI services in any of these application categories are contractually prohibited from doing so under these terms, regardless of jurisdiction.
⚠ Use of Microsoft AI services in prohibited categories constitutes a breach of these terms and may trigger account enforcement actions including service termination as stated in Section 4.b.iv
Cross-platform context
See how other platforms handle AI Usage Restrictions: Prohibited High-Risk Applications and similar clauses.
Compare across platforms →"You agree that you will not use the AI services, including those that make decisions, or take actions, autonomously or with varying levels of human intervention: 1. To make decisions or take actions without appropriate human oversight that may have a consequential impact on any person's legal position, financial position, life opportunities, employment opportunities, or human rights, or may result in physical or psychological harm to a person or group of persons; 2. To deceive or intentionally misinform (e.g., false advertising), or deploy subliminal techniques (e.g., visual, auditory, or other signals beyond a normal person's range of perception) with the intent to manipulate or distort the behavior of a person in a way that causes harm; 4. For social scoring or predictive profiling that would lead to discriminatory, unfair, biased, detrimental, unfavorable, or harmful treatment of certain persons or groups of persons; 6. Based on their biometric data, to categorize people or to deduce or infer their race, political opinions, trade union membership, religious or philosophical beliefs, or sex life or sexual orientation; 7. To create or expand facial recognition databases through the untargeted scraping of facial images from the internet or CCTV footage.Excerpt from Microsoft Copilot's Terms of Service
1) REGULATORY LANDSCAPE: These prohibitions directly engage the EU AI Act's list of prohibited AI practices, particularly Articles 5(1)(a) through 5(1)(h) covering manipulation, exploitation of vulnerabilities, social scoring, real-time biometric identification, and predictive profiling.
Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.
Get the research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.
These nine prohibited use categories in Section 14.s.ix closely track prohibited AI practices under the EU AI Act and represent contractual restrictions that apply to all users of covered AI services globally. Organizations deploying Microsoft Copilot or other AI services in any of these application categories are contractually prohibited from doing so under these terms, regardless of jurisdiction.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Microsoft Copilot.