Prof. Pekka Santtila, PhD – Professor of Psychology at New York University (NYU) Shanghai and Global Network Professor at NYU – is a leading expert in legal psychology with over 300 scientific publications in this field. He actively collaborates with police authorities, testifies as an expert witness in court and advises on investigations into child sexual abuse. In addition, Professor Santtila researches applications of artificial intelligence in the legal system – from assisting in investigative interviews to analysing jury decision-making.

Last week, he gave a lecture on the future of questioning in the courtroom at the annual conference of the European Association for Psychology and Law (EAPL 2025) in Vilnius, which was attended by around 250 participants from 34 countries. The event was organised by the Applied Psychology Research Laboratory at Mykolas Romeris University (MRU).

“Unfortunately, traditional training does not always lead to long-term results,“ notes Prof. Dr. P. Santtila.

To assist investigators in preparing more effectively, he and his teams developed an innovative tool – avatars. These virtual characters (originally designed as children, now also available as adults) enable practice in realistic environments. Such training helps to formulate better questions, avoid suggestiveness, and adapt to different situations. “Avatars provide real-time feedback, enable on-the-spot corrections and strengthen skills that can later be applied in actual investigations,“ explains the professor.

Another area of innovation is large language models (LLMs): AI technologies that can assist researchers during interviews by suggesting questions, identifying potential bias, or even monitoring emotional tone. “We are moving towards a hybrid interview model in which humans and AI work together. I expect that in the near future, researchers will wear VR headsets that help them maintain high quality even under stress,“ he predicts.

Studies show that AI can not only facilitate interviews, but also help reduce bias in statements. In an experiment with children aged 6 to 8, for example, AI-assisted interviews were more accurate: artificial intelligence asked fewer unnecessary questions and collected less incorrect information.

Another interesting approach currently being developed involves individualised avatars whose behaviour reflects the profile of a specific suspect. An investigator could prepare for a specific interrogation in advance by training with a virtual character. “This could fundamentally change the way interrogations are prepared and make them truly personalised,“ says Prof. Dr. Santtila.

However, innovation requires responsibility. With regard to artificial intelligence, the professor emphasises that maintaining human control is of utmost importance. “AI must serve as an assistant, not a decision-maker. Humans must monitor the process and take responsibility,“ he stresses. It is also important to ensure that AI remains transparent and unbiased so that it is always clear how and why a particular decision was made.

Technological advances in survey research may seem like a futuristic scenario, but according to the professor, they are already happening. The most important thing is to use innovations responsibly, in compliance with ethical principles and scientific standards, and for the benefit of people – especially those who are most at risk.