This 28^th-30th July, Lancaster University’s Forensic linguistics, Cybersecurity and Technology Research (FACTOR) team, based in the Department of Linguistics and English Language, ran their three-day Summer School. The event took 77 delegates on a journey into the worlds of forensic linguistics (FL), forensic speech science (FSS), and the (mis)use of AI, and helped to raise awareness about a new unique course, MSc Forensic Linguistics and Speech Science.

The summer school attracted a broad mix of University staff, students, alumni, businesses, members of the public sector and the general public, learning from experts at the very forefront of their fields. The Department of Linguistics and English Language is consistently ranked third in the world (QS2024, QS2025). Lancaster University is one of only a handful of universities nationally to be identified by the National Cyber Security Centre as an Academic Centre of Excellence in Cyber Security Research and Cyber Security Education.

Speakers included Professor Claire Hardaker (Forensic Linguistics, Director of FACTOR), Dr Georgina Brown (Forensic Speech Science, Data Science Institute (DSI) Deputy Theme Lead for Responsible Security), Dr Isobelle Clarke (Forensic Linguistics, Security & Protection Science), and Dr Justin Lo (Forensic Phonetics, Security & Protection Science), supported by their team of PhD students. The FACTOR team are also members of Security Lancaster, Security and Protection Science, and the DSI, which part-funded the Summer School.

Delegates learnt how voice identification is important for intelligence purposes and in the criminal justice system, for example hoax calls, scam calls to insurance companies, CCTV and surveillance recordings. They explored how to analyse voices for features for identification, such as tone, quality, accent, intonation, pitch, learned about what characteristics make us trust some people more than others, for example face shape, occupation and accent, and to notice the linguistic habits people unconsciously fall into. They then applied this knowledge in an authorship analysis task- identifying whether samples of text were all likely to be written by a suspected catfisher in a romance scam fraud case study.

The speakers shared some of the latest research exploring how well AI is currently able to replicate our voices, and delegates tried detecting whether voice samples were AI-Generated or not, or “Bot or Not.” They also got to test their analytical skills on the two main types of technology that produce AI- Generated speech- Text-to Speech (TTS) systems and Voice Conversion (VC) systems, and learnt how to look for tells/features of each.

Through detailed case studies of real crimes, they were shown how forensic linguistic investigations actually take place, how analysing enormous datasets of text can help with authorship analysis casework, and how these findings can be used as part of a body of evidence to investigate terrorists and murderers. They were also shown how AI is being used to waste scammers’ time, for example using AI for scambaiting, and how AI can help protect vulnerable people by anonymising witness statements.

On the last day they played the role of investigative expert with a high-profile fraud case to solve, combining their new skills in AI detection, voice comparison, and authorship analysis, to try and identify the culprit in a bank scam based on a real case in Hong Kong in 2024 where an elaborate deep-fake video call was used to steal £20m from a bank.

Feedback on the event was extremely positive, with attendees commenting on the “engaging, informative, helpful and friendly staff,” “great content and development pace,” describing it as “super,” “interesting,” “thought-provoking,” “inspirational,” “10/10”, a “wonderful three days” which “satisfied beginner(s) with little knowledge,” those who “loved being able to dip my toe back into linguistics world, and for other it helped “emphasise (their) love for linguistics!”

There are wide applications for forensic linguistics. Professor Claire Hardaker concluded, “One of our highest priorities in FACTOR is using our research for real-world good, and given the swift rise of AI-powered crime, we’re working hard to help ordinary people understand the new reality we’re heading into. From romance scams to inside-job bank heists, no matter who you are, AI is going to have an increasing array of impacts on your life in various ways, and as far as possible, we’d like to ensure that the positives of this emerging technology outweigh the negatives.”

The FACTOR Team intends to run the Summer School again in 2026 and are also open to developing sector specific and bespoke training sessions for industry and public services, for example, banking and the police.

Contact factor@lancaster.ac.uk or see FACTOR | Forensic linguistics, Cybersecurity and Technology Research and sign up for blog alerts.

To find out more about the MSc Forensic Linguistics and Speech Science, see Forensic Linguistics and Speech Science MSc - Lancaster University

Play the real Bot or Nots

· Music Edition: can you tell your Billie Eilish from your Billie AI-lish?

· Speech Edition: can you your doting aunt from your digital agent?

· Text Edition: can you spot the scam, AI-generated hotel reviews?