This report is from last year and I don’t think I’ve written about it yet, and The Wall Street Journal last week ran an article from one of the report’s authors that resurfaced it, so I wanted to either write about it again or write about for the first time because it’s interesting that artificial intelligence may help consumers get better results when complaining, especially with regulators like the Consumer Financial Protection Bureau.
Here is the key takeaway: consumers who used large language models to draft complaints were significantly more likely to receive relief from financial institutions. That alone should get your attention if you are thinking about how consumers are evolving and how AI is changing the dynamics of dispute resolution.
What the study looked at: Researchers analyzed more than 1 million complaints submitted to the CFPB between 2015 and 2024 .
- They identified complaints likely written with AI tools such as ChatGPT
- They tracked adoption trends before and after ChatGPT’s release in late 2022
- They compared outcomes between AI-assisted complaints and human-written complaints
The results were clear. After ChatGPT launched, AI usage in complaints rose quickly, going from near zero to nearly 10% of complaints by early 2024 .
The outcome difference is what matters: Consumers using AI had a materially higher success rate.
- AI-assisted complaints: 49.3% received relief
- Human-written complaints: 39.9% received relief
That is roughly a 10 percentage point lift, which held even after controlling for issue type, product, and other variables. Researchers also used statistical methods and controlled experiments to test causality, and the results held up.
Why this is happening: The advantage is not about new facts. It is about better communication.
The study found that AI improves:
- Clarity
- Structure
- Professional tone
- Coherence and persuasiveness
In controlled experiments, the exact same complaint content performed better when rewritten by AI, suggesting that presentation alone can influence outcomes.
What this means: There are a few implications that are hard to ignore:
- Consumers are getting more sophisticated
AI is effectively acting as a writing assistant, helping consumers present stronger, more polished cases. - Your frontline teams are now dealing with “optimized” complaints
These are clearer, more structured, and potentially harder to dismiss or misunderstand. - There may be a perception risk
If better-written complaints get better outcomes, it raises questions about consistency and fairness. - AI is leveling the playing field
The study notes that adoption was initially higher among consumers with language barriers, suggesting AI can help bridge communication gaps.
