Only 6% of the 66 academic librarians surveyed currently use AI-enabled alert services, while 24% expect to begin using them or increase their use over the next 12 months. The report examines which tools librarians use to track new research and professional developments, how they receive alerts, and what concerns shape their decisions.
Three of the seven library deans and directors surveyed reported current use. Interest in expanding use was strongest among respondents working in electronic resources, licensing, or acquisitions. Five of eight in that group expected to start or expand their use over the next year, although none reported having used or tested an AI alert service. Across the full sample, 42% were unsure about their plans or did not answer the question.
Alerts from established research sources such as Google Scholar, Scopus, Web of Science, and PubMed, used in combination with AI, were the most commonly cited category, at 12%. Microsoft 365 Copilot or Power Automate followed at 9%, and ChatGPT Tasks or another scheduled ChatGPT workflow at 8%. New publications and other research outputs topped the list of monitored topics, followed by citations. Email digests were the most common delivery method.
In written responses, some librarians described using alerts to discover relevant papers or follow news tied to university priorities. When asked what advice they would give peers, 14 of the 23 respondents who commented discouraged use or expressed a negative view. They cited unreliable results, ethical concerns, and environmental costs. Librarians who offered practical advice suggested setting up alerts for a clear purpose, routing them to a dedicated inbox, and checking the results before relying on them.
The report presents results by job responsibility, institution type, enrollment, and other characteristics. Some subgroup comparisons reflect only a few respondents.