Study: AI Scripting Conversations; Qualitative Insights

According to The Next Web on July 27, 2026, young adults are increasingly using AI chatbots to script in-person conversations, and psychologists warn this trend promotes what they call an "epidemic of self-mistrust."

• A July 27, 2026 report in The Next Web cites a study finding AI advice made people three times less accurate but twice as confident when assessing correctness.
• The July 27, 2026 article documents multiple 20-year-old users who rely on chatbots for dating and conflict-preparation, including Niharika Abbaraju and Kalin Lincoln.
• Clinical psychologists quoted by The Next Web on July 27, 2026 describe this pattern as an "epidemic of self-mistrust, " and relationship coach Alexandra Friedman recommends capping AI rehearsal to once a week.

Answer: The Next Web reported on July 27, 2026 that a wave of young adults are using chatbots to draft both text and in-person conversational lines, based on interviews and linked studies.

According to The Next Web on July 27, 2026, the reporting draws on first-person accounts such as 20-year-old cancer researcher Niharika Abbaraju who uses ChatGPT at work and as a personal conversation coach off the clock.

According to The Next Web on July 27, 2026, clinicians and researchers identify a behavioral pattern where people defer to AI phrasing, and Wharton researchers call the broader decision-making variant "cognitive surrender, " as reported by The Next Web on July 27, 2026.

According to The Next Web on July 27, 2026, the phenomenon shows up in practice as edited grammar on dates, AI-sourced summaries to cover unread books, and chatbot-prepared calming scripts before difficult conversations.

Answer: Researchers must treat AI-assisted social scripting as a distinct codebook category to avoid misattributing polished responses to personal skill, as argued by The Next Web on July 27, 2026.

Researchers should add direct probes to interviews, for example: "Did you use any app or chatbot to draft or rehearse that reply? ", a probe recommended after the patterns described in The Next Web on July 27, 2026.

Researchers should triangulate reported behavior with interaction traces, and The Next Web on July 27, 2026 indicates that observable cues include sudden grammar shifts and unnatural phrasing during live conversations.

Researchers should code for "cognitive surrender, " the label used by Wharton researchers and reported by The Next Web on July 27, 2026, to capture deferral to AI even when the AI may be wrong.

Qualitative researchers should treat AI scripting conversations as a distinct behavior to measure, code, and report, as documented by The Next Web on July 27, 2026.

Researchers can start by adding direct probes, creating a "cognitive surrender" code, and using automated transcript analysis to find stylistic markers across interviews.

Evidano speeds these steps by ingesting transcripts, surface patterns, and producing segment comparisons; for more detail see our features page.

If you want to try these techniques on your own data, Try Evidano for free.