No class action lawsuit currently exists targeting LinkedIn for scraping member content to train Microsoft Copilot AI. However, the situation is more detailed than the premise suggests: LinkedIn and Microsoft have announced plans to scrape member profile data from users in certain regions (EU, EEA, Switzerland, Canada, and Hong Kong) to train their own AI models as of October 27, 2025.
Meanwhile, LinkedIn has been the plaintiff in multiple lawsuits against unauthorized data scrapers who illicitly harvested member information. Additionally, a separate class action targets Microsoft, GitHub, and OpenAI over Copilot’s training on publicly available GitHub code without proper attribution. This article clarifies the actual litigation landscape, explains LinkedIn’s announced data practices, and explores what these developments mean for member privacy and potential legal exposure.
Table of Contents
- Did LinkedIn Actually Sue Over Copilot Training?
- The Real Litigation: LinkedIn as Plaintiff Against Data Scrapers
- The GitHub Copilot Class Action: The Real AI Training Litigation
- LinkedIn’s Announced Data Use: What Members Need to Know
- Privacy Risks and Your Rights
- What Makes This Different from Other Data Scraping
- The Future of AI Training and Data Rights
Did LinkedIn Actually Sue Over Copilot Training?
No—the specific class action described in the title does not exist. What does exist is LinkedIn’s own announcement that it will scrape member profile data for AI training purposes, sharing data with Microsoft subsidiaries. This is different from being sued for the practice.
The confusion may stem from legitimate privacy concerns about tech companies using personal data for AI without explicit, granular consent. LinkedIn’s announcement explicitly named the regions where scraping would occur (EU, EEA, Switzerland, Canada, and Hong Kong), suggesting regional data protection laws influenced the decision. In regions not listed, LinkedIn’s terms of service may already permit such use, though members typically discover these practices only after announcement or through privacy advocates.
The Real Litigation: LinkedIn as Plaintiff Against Data Scrapers
The actual legal battles involving LinkedIn and data scraping paint a different picture. LinkedIn and Microsoft filed a lawsuit against ProAPIs, a company operating millions of fake accounts to scrape member data. The case reached an “agreement in principle” settlement in U.S. District Court for the Northern District of California, though specific damages were not disclosed publicly.
Similarly, LinkedIn sued Proxycurl, a data enrichment API company, alleging that Proxycurl operated “hundreds of thousands” of fake accounts to harvest data from millions of users. This litigation forced Proxycurl to shut down. The irony is striking: while LinkedIn aggressively pursues unauthorized scrapers through litigation, the company simultaneously announced its own member data scraping for AI purposes. The difference, LinkedIn would argue, is authorization—their terms of service presumably permit the practice, whereas scrapers violate the law.
The GitHub Copilot Class Action: The Real AI Training Litigation
The class action most directly related to AI training without proper attribution involves GitHub Copilot, not LinkedIn. Developers filed suit on November 3, 2022, against Microsoft, GitHub, and OpenAI, alleging that Copilot was trained on billions of lines of public GitHub code without proper attribution or licensing compliance.
Copyright infringement claims were dismissed in May 2023, but DMCA (Digital Millennium Copyright Act) anti-circumvention claims and breach of contract claims survived and remain active. This litigation demonstrates that courts are wrestling with whether training AI models on publicly available data without attribution or compensation constitutes copyright infringement or breach of terms of service. A similar argument could theoretically apply to LinkedIn’s profile scraping, though LinkedIn’s terms of service would need to be examined carefully to determine whether members have already consented.
LinkedIn’s Announced Data Use: What Members Need to Know
On October 27, 2025, LinkedIn publicly announced that it would begin scraping profile data from members in specific regions to train its own AI models, with data shared to Microsoft subsidiaries. This announcement came with a critical caveat: it applied only to the EU, EEA, Switzerland, Canada, and Hong Kong—regions with strict data protection regulations like GDPR. Members in other regions (including the United States, for the most part) were not granted the same opt-out opportunity, suggesting LinkedIn either believes existing terms of service already permit the practice in those jurisdictions or that legal risk is lower.
The distinction matters: in jurisdictions with stronger data protection laws, LinkedIn was forced to offer transparency and choice. In others, the company proceeded under existing terms. Members in affected regions could theoretically challenge this in court if they can demonstrate that the terms of service did not clearly authorize AI training specifically.
Privacy Risks and Your Rights
The practice of scraping profile data for AI training carries several privacy risks that members should understand. First, even “anonymized” training data can sometimes be reverse-engineered to identify individuals, especially if combined with other datasets. Second, training AI on LinkedIn profiles means your professional history, connections, skills, and endorsements are being used to improve a commercial product you may not have explicitly consented to.
Third, once data is used to train an AI model, it becomes nearly impossible to truly delete—the model encodes patterns from that data permanently. However, in jurisdictions with strong data protection laws, members may have limited legal rights to demand deletion or compensation, though such challenges are still being tested in courts. If you are concerned about your data being used, reviewing LinkedIn’s updated privacy policy and terms of service is essential, and members in covered regions should have received notification about opting out.
What Makes This Different from Other Data Scraping
LinkedIn’s announced AI training program differs from unauthorized scraping in one key way: authorization. When Proxycurl scraped LinkedIn data, it violated the Computer Fraud and Abuse Act and LinkedIn’s terms of service. When LinkedIn scrapes its own members’ data under its own terms of service, it argues the practice is authorized and lawful.
The critical question, however, is whether members actually understood and agreed to such use when they created their profiles. Many terms of service are written broadly (“we may use your data to improve our services”) and updated frequently, making true informed consent questionable. This is where future litigation may emerge: not from LinkedIn scraping for AI, but from members arguing that the terms of service were unconscionable, misleading, or violated privacy laws by not obtaining specific consent for AI training.
The Future of AI Training and Data Rights
The landscape around AI training and data rights is evolving rapidly. The GitHub Copilot litigation will likely establish precedents about whether training AI on publicly available data without attribution is fair use or copyright infringement. LinkedIn’s announced scraping practices may face challenges in court, particularly if members in non-covered regions file class actions arguing that the terms of service did not adequately disclose AI training purposes.
Regulators in the EU and elsewhere are also developing AI-specific frameworks that may impose stricter requirements on companies using training data. The broader trend suggests that future AI development will face greater scrutiny around consent, transparency, and compensation for data use. Members should monitor developments in AI litigation and privacy law, as court decisions or regulatory changes could affect their rights and options.