Original title: Informational Help-Seeking on Reddit Did Not Decline After ChatGPT
Hazem Ibrahim, Yasir Zaki · September 2026
Did people stop asking others for advice online once ChatGPT arrived? Researchers wanted to find out. Earlier studies disagreed because they did not compare help forums to similar groups where artificial intelligence cannot easily replace people, over the exact same months. To test this, the researchers compared post numbers across 26 Reddit advice communities with 90 hobby communities over six calendar months before and after ChatGPT launched. They also ran this test across 66 earlier dates before ChatGPT existed. They found that informational help-seeking did not drop. Their findings rule out any posting decrease larger than 3.4%, which is far smaller than past estimates of 8% to 25% drops. Bot detectors also showed that AI did not replace human activity. AI posts rose just 2-3 percentage points more in advice groups, and replies showed no rise at all. Humans still ask humans for help.
Imagine you run a help desk where people call for advice. You worry that a new automated assistant might make people stop calling. You check the call logs and see no drop. But maybe the automated assistant is secretly answering calls, so you check recordings and find only a few calls are automated. In this paper, the researchers did the same for Reddit: they counted posts in advice communities, found no decline, and checked for AI-written posts, finding only a tiny increase.
This matters because it shows that, contrary to some fears, generative AI has not yet replaced human advice-seeking in online communities. For people who rely on Reddit for practical information, the community remains active. For researchers, it highlights how failing to compare affected communities against similar unaffected ones can produce misleading estimates of AI's impact.
The study only covers Reddit, not other platforms. It examines only six months after ChatGPT's launch, so longer-term effects remain unknown. AI-text detectors have limited accuracy, and undetected AI-written posts could still exist. The analysis cannot fully rule out that AI-generated content influenced posting behavior in ways not captured by post counts or detection.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Informational Help-Seeking on Reddit Did Not Decline After ChatGPT — Hazem Ibrahim, Yasir Zaki
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.