Truace
Subscribe
70/100

Moderate evidence

Trace: a gain and a problemPublished recordBusinessThe Trace

Full story
More
Export record JSONEmailLinkedInX

Empowered generative artificial intelligence in retail service recovery: Recovery justice, customer effort reduction, trust restoration, and purchase behaviour

Truvace record
TRV-2026-2585raw data (JSON)
Version
v1Published
First published
Evidence type
Randomized controlled trial
Sources
1 sourcePeer-reviewed
sha256
d4f998b6fd62…check it

Retailers increasingly deploy generative artificial intelligence (GenAI) to handle complaints, yet evidence remains limited on when these agents restore purchasing rather than merely improve immediate evaluations. We develop a configuration model in which recovery agent identity (GenAI versus human) is separated from recovery agent empowerment, defined as authority to complete a remedy during the encounter. Across a qualitative study, two pretests, two randomised experiments, and two secondary-data studies in Un…

59Gain side

How the two sides scoreClose call

58Problem side

The gain

In experiments and UK retail field studies, generative AI complaint agents that could issue refunds, replacements, delivery upgrades or credits improved perceived justice, cut customer effort, restored trust and produced the strongest purchase outcomes.

The problem

Generative AI complaint agents without visible authority to complete a remedy were not better than human agents, and the study found customers with strong relationships penalised low-empowerment AI more.

A fixed formula sets both scores from the evidence and its sources. Reader votes don’t move them.

What the scores mean

G is the gain side, out of 100. It goes up when the gain is specific, when it measurably helped people and when the sources behind it are strong.

P is the problem side, out of 100. It goes up when the problem is specific, when it measurably hurt people and when the sources behind it are strong.

What this doesn’t fix

Failure severity weakened the advantage of empowered generative AI, so the benefit does not hold equally for severe or data-constrained complaints.

Sources

1 on file
  1. Journal of Retailing and Consumer Servicesdoi:10.1016/j.jretconser.2026.105042
    Peer-reviewed2026-08-18

Cite this record

v1

Truvace Impact Record TRV-2026-2585, v1: “Empowered generative artificial intelligence in retail service recovery: Recovery justice, customer effort reduction, trust restoration, and purchase behaviour.” Truvace, 2026-11-12. /record/TRV-2026-2585 (accessed on the day you copy it). sha256 d4f998b6fd628076…

Change history

1 version

Every change made to this record since it was first published, and why. Nothing has changed since it went up.

  1. Publishedv1d4f998b6fd62…

    Certified into the record

Check this record hasn’t been altered

Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-2585 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.