Recommendation SystemConsumer FacingVerified

Reviewed and published by trentmaziarz, March 18, 2026. Discovered and drafted by our automated research pipeline.

Adobe Target uses AI to automatically personalize what individual visitors see on a website or app — testing different combinations of content, offers, and designs and gradually directing more visitors to whichever version performs best for each person. The AI Recommendations feature suggests products or content based on each visitor's behavior. These features have been available since approximately 2016–2018.

Details

Adobe Target Premium offers four AI-powered modes built on a random forest machine learning algorithm: Auto-Allocate shifts traffic toward better-performing variations using a multi-armed bandit approach; Auto-Target personalizes the experience for each visitor individually based on browsing history, location, and other signals; Automated Personalization matches specific offer combinations to individuals; and Recommendations applies collaborative filtering to suggest products or content based on what similar visitors viewed or purchased. The Recommendations algorithm was adapted from machine learning models originally developed for Behance's content discovery feed.

Products affected

Adobe Target

Sources & Evidence

Company Disclosure

Company DisclosureVerified

Experimentation and optimization

Adobe Business·Jan 2024

https://business.adobe.com/products/target/experimentation-optimization.html

Cite this record

Trace Foundation. (2026). Adobe: Adobe Target uses AI to automatically personalize what individual visitors see on a website or app — testing different combinations of content, offers, and designs and gradually directing more visitors to whichever version performs best for each person. The AI Recommendations feature suggests products or content based on each visitor's behavior. These features have been available since approximately 2016–2018 (data as of 2026-03-18) [Data set record]. AI Trace. https://www.aitrace.org/r/practice/fa4b8569-5eac-45dc-be11-4ca09c79de76. Accessed October 13, 2026.

Stable link: https://www.aitrace.org/r/practice/fa4b8569-5eac-45dc-be11-4ca09c79de76
Data as of: March 18, 2026
Last verified: February 28, 2026

How to cite AI Trace

Other practices by Adobe

OtherIn June 2024, Adobe pushed updated Terms of Use to all Creative Cloud users that required mandatory acceptance before they could continue using their software. The language alarmed many users and creative professionals who believed Adobe was claiming the right to use their personal work to train AI models. Adobe published clarifications and then released fully revised terms later that month.OtherAdobe InDesign version 21.2, released in January 2026, added an AI feature that automatically generates alt text — written descriptions of images used by screen readers to assist people who are blind or have low vision. The feature is on by default and has received mixed reactions from users about its accuracy.OtherAdobe has maintained a formal set of AI ethics principles since 2019, built around three pillars: Accountability, Responsibility, and Transparency. Every product team developing an AI feature must complete an internal AI Ethics Impact Assessment before launch. Adobe updated these principles in October 2024.

Have evidence about Adobe's AI practices? Submit a report.

Report a Sighting →