Millisecond Forums

+x

 

Dave - 2/20/2026

+x

 

EN - 2/20/2026

Hi,
I was wondering whether there is a way to prevent participants from pasting text into an <openended> response box, particularly when using Inquisit Web.

This question comes from recent concerns about participants’ (mis)use of LLMs or other AI‑assisted tools to generate responses.

Relatedly, does Inquisit Web offer any functionality that could help reduce the likelihood that AI agents or AI‑assisted technologies interfere with a study?

Finally, are there any built‑in ways to detect the use or presence of such tools—for example, beyond recording and inspecting response timestamps throughout the task?

Any insights or experiences would be greatly appreciated. Thanks in advance!

E

I'm afraid the answer to all your questions is no. There is no way to prevent pasting text. Nor is it possible to detect presence or use of AI tools. Doing so would be error-prone, extremely intrusive, and -- beyond that -- not technically feasible on various platforms (Android, iOS), which deliberately and reasonably limit the access apps have to the system at large and what other apps exist or are doing on the given system.

Thanks for the reply, Dave

+x

 

EN - 2/20/2026

Hi,
I was wondering whether there is a way to prevent participants from pasting text into an <openended> response box, particularly when using Inquisit Web.

This question comes from recent concerns about participants’ (mis)use of LLMs or other AI‑assisted tools to generate responses.

Relatedly, does Inquisit Web offer any functionality that could help reduce the likelihood that AI agents or AI‑assisted technologies interfere with a study?

Finally, are there any built‑in ways to detect the use or presence of such tools—for example, beyond recording and inspecting response timestamps throughout the task?

Any insights or experiences would be greatly appreciated. Thanks in advance!

E

I'm afraid the answer to all your questions is no. There is no way to prevent pasting text. Your best bet for openended elements may be to log and inspect the firstcharlatency and lastcharlatency properties, which reflect the time of first typed and last typed character respectively. If, for a lengthy response, the difference is too short for the response to have been typed, the likelihood is high that it was pasted.
https://www.millisecond.com/support/docs/current/html/language/properties/firstcharlatency.htm
https://www.millisecond.com/support/docs/current/html/language/properties/lastcharlatency.htm

Nor is it possible to detect presence or use of AI tools. Doing so would be error-prone, extremely intrusive, and -- beyond that -- not technically feasible on various platforms (Android, iOS), which deliberately and reasonably limit the access apps have to the system at large and what other apps exist or are doing on the given system.

Hi,
I was wondering whether there is a way to prevent participants from pasting text into an <openended> response box, particularly when using Inquisit Web.

This question comes from recent concerns about participants’ (mis)use of LLMs or other AI‑assisted tools to generate responses.

Relatedly, does Inquisit Web offer any functionality that could help reduce the likelihood that AI agents or AI‑assisted technologies interfere with a study?

Finally, are there any built‑in ways to detect the use or presence of such tools—for example, beyond recording and inspecting response timestamps throughout the task?

Any insights or experiences would be greatly appreciated. Thanks in advance!

E