The Adverse Outcome Pathway (AOP) development and assessment process is very time-consuming, as an enormous number of literature references need to be read, quality-checked, and assessed. Relevant content needs to be extracted, and the selected knowledge must be translated into KEs, KERs, and assembled into an AOP, and the AOP needs to be assessed and reviewed. The sheer effort required for humans to accomplish these tasks is often prohibitive, causing scientists to shy away from even considering developing an AOP. While the number of AOPs in the AOP-Wiki is still increasing and the AOP concept is now an established scientific concept, there is a risk that a critical mass of AOPs (which would herald a true paradigm shift in science) will not be reached unless a new approach supports a significant growth in AOP numbers.The main problem of the AOP Framework is its linear scaling: When 2 AOPs require 10 scientists (and reviewers), then 200 AOPs require 1000 scientists (and reviewers) – and this is not sustainable.It can be assumed that AI can support many aspects of the AOP development, assessment and review process. AI4AOP will be an effort to explore the possibilities of using AI to support the key processes in the AOP domain. Ideally, a group of like-minded individuals will work together to leverage the power of AI to overcome the bottleneck of specialized expertise in creating a critical mass of AOPs.The group will form, will select topics to be treated and will agree on desired goals and deliverables. This slide deck is an introduction to the effort. It was first shown on 2024-10-30 during a meeting of the Society for the Advancement (SAAOP) Knowledgebase Interest Group (SKIG).