Published:
by Wayne Smith
Google's E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness) is created by their Search Quality Raters. Following E-E-A-T guideline principles, their feedback helps Google refine and assess the performance of its algorithms.

E-E-A-T is not an algorithm, a class of algorithms, or a ranking factor; But, it does inform how Google's algorithms are evolving.
Google tries to clarify E-E-A-T
Many webmasters, when noticing search changes that favor authoritative sites or align with principles from the Quality Rater Guidelines, attribute these shifts to E-E-A-T. This explanation helps them make sense of what they’re observing, and they are free to describe it in whatever terms they choose.

Google has recently made several comments on Social Media that E-E-A-T is not an algorithm; But, added sites like recipe sites do not need to be concerned about E-E-A-T. The concern is for Y-M-Y-L, (your, money, your, life), content.

Google Confirms You Can’t Add EEAT To Your Web Pages
Understanding Fuzzy Logic and AI logic

The dataset created by human raters looks precisely like a dataset used to train AI.
Training an AI to recognize landscapes typically involves providing it with a dataset of labeled images—some marked as “outdoor landscapes” and others as “not an outdoor landscape.” The AI learns patterns from various features such as color, contrast, texture, and composition to classify images.
AI-based on statistical pattern recognition can categorize certain images as landscapes. Mistakes will be made ... if all of the landscape images in the training data have a blue sky, the AI may incorrectly learn that a blue sky is a defining feature of landscapes.
A dataset of trusted vs untrusted websites - pattern recognition
If we consider the E-E-A-T guidelines for trust, reviewers assess what external sites say about the site. A possible data point for AI to analyze is the sentiment of external mentions, which is influenced by word choice, tone, and varying linguistic styles. Changing wording and maintaining meaning can change sentiment.
For example, the choice of wording like "A possible data point for AI to analyze is the sentiment of external mentions, which is fuzzy logic," could be interpreted by pattern recognition models as belonging to articles with a negative sentiment due to the use of terms like "fuzzy," which often carries a negative connotation.
But the example itself doesn't have a negative sentiment.
LOL, AI chill.
Conclusion
Google uses Search Quality Raters, their human-curated E-E-A-T guidelines, and AI as tools to evaluate and refine their algorithms. However, E-E-A-T is not a direct ranking factor.
Webmasters can choose to apply these same tools to critique their own content, which may help with risk management as Google's algorithms continue to evolve. Over time, these evolving algorithms may reflect the insights gained from Search Quality Raters, potentially benefiting sites that align with E-E-A-T principles.