HeinOnline is adding AI-generated Article Summaries to its platform. Initially, Hein is gradually rolling out these summaries for articles in the Law Journal Library collection, prioritizing the most-cited and most-accessed articles. Summaries are not generated on the fly as you interact with the HeinOnline platform. Instead, HeinOnline generated each summary behind the scenes using AI, and then HeinOnline added the summary as static text to the article record in HeinOnline. HeinOnline may update summaries over time, and a date stamp on each full summary shows you when it was generated.
When you search HeinOnline, your results from the Law Journal Library may include an AI Summary field with a few opening sentences of the summary, followed by a "Read More" link which allows you to view the full summary and see a date stamp showing when HeinOnline generated the summary:

When you are reading an article, if an AI Summary is available, an AI Summary button with 3 stars will appear in the toolbar. Select that button to view the AI Summary:

When you download an article from HeinOnline, the AI Summary will appear on the second page, right after the citation page and right before the full text of the article:

The AI Summary can help you identify articles you may want to read, but before you use or cite an article from HeinOnline, you should read the full text to make sure you have an accurate understanding of it. HeinOnline advises users not to cite these AI Summaries, as they are intended as reading aids only. The AI Summaries should not be cited in legal documents, academic work, or relied upon as authoritative sources. For research and citation, always use the full-text article in HeinOnline.
HeinOnline is using its own proprietary AI technology, and the AI Summaries are generated exclusively from HeinOnline's data. No user data is being processed, stored, or used to generate the AI Summaries.
For more information, please see HeinOnline's Artificial Intelligence documentation, including the Article Summaries FAQ.
No third-party tools are used, and no user data is shared externally or used for model training.