Source URL: https://www.theregister.com/2025/02/12/thomson_reuters_wins_ai_copyright/
Source: The Register
Title: A win at last: Big-time blow to AI world in training data copyright scrap
Feedly Summary: You gotta fight … for your Reuters … to party
Thomson Reuters has won a partial summary judgment in a copyright case against shuttered AI firm Ross Intelligence, a decision that disallows fair use as a defense for training models on proprietary data without permission.…
AI Summary and Description: Yes
**Summary:** The text reports on a significant legal decision in the copyright case between Thomson Reuters and AI firm Ross Intelligence, disallowing the use of fair use as a defense for training AI models on proprietary data. This sets a precedent impacting the ongoing discourse around AI copyright issues.
**Detailed Description:**
The legal ruling represents a pivotal moment in the intersection of artificial intelligence and copyright law, underscoring potential liabilities for AI firms that rely on proprietary content without proper licensing. Key points include:
– **Case Background:**
– Thomson Reuters brought a lawsuit against Ross Intelligence for using its Westlaw editorial content without permission to train AI models, claiming copyright infringement.
– Ross Intelligence attempted to source legal content through bulk memos but was eventually sued by Thomson Reuters, which argued that such use violated copyright protections.
– **Court’s Decision:**
– Federal district Judge Stephanos Bibas ruled against Ross Intelligence, stating that its use of Thomson Reuters’ content did not qualify as “fair use.”
– The decision is particularly notable as it focuses on training AI model inputs rather than the outputs produced by AI.
– **Implications for AI Firms:**
– This ruling could embolden other plaintiffs involved in the 38 pending AI-related copyright claims in US courts.
– Experts believe it may influence future court cases related to generative AI, prompting legal parties to cite this ruling when seeking determinations on copyright infringement and fair use.
– **Legal Analysis:**
– The judge articulated that Ross’s actions were not transformative and directly competed with Thomson Reuters’ offerings, thereby failing the fair use criteria.
– The court’s interpretation of copyright applies specifically to legal content, suggesting a specialized analysis for legal information and AI.
– **Expert Opinions:**
– Legal scholars contend that the ruling may not only impact Ross Intelligence but will be a critical reference point for future disputes involving AI and proprietary data.
– With implications for potential appeals and the likelihood that these issues will reach the Supreme Court, the discourse on fair use for AI training continues to evolve.
With technology continuously advancing, companies operating in the AI space need to prioritize understanding copyright law and implementing robust compliance strategies to mitigate risks associated with proprietary data usage.