Exploring Digital Profiling

'Digital Profiling and its Applications in Research and Practice'

Date: Tues 18 – Wed 19 November, 2025
Location: Manchester City Centre
Speakers: Dr Joanne Hinds, Dr Heather Shaw, and Professor Oli Buckley.

Early career researchers gathered in Manchester for the NABS+ Network Plus event: Digital Profiling and its Applications in Research and Practice. Over two days, leading researchers, including Dr Joanne Hinds, Dr Heather Shaw, and Professor Oli Buckley, who introduced and explored the science, ethics, and future of digital profiling with participants. 

Dr Joanne Hinds opened the event with a thought-provoking overview of digital profiling, describing it as the process of collecting and interpreting digital traces to infer aspects of identity, personality, and behaviour. She challenged the common assumption that profiling relies mainly on browsing history, demonstrating instead how contemporary platforms such as TikTok, Spotify, and Amazon draw on rich behavioural signals to predict significant life events, like relationship breakdowns and pregnancy. Jo introduced personality assessment through the Big Five framework and explored key theoretical models used to interpret digital behaviour. These included the Self–Other Knowledge Asymmetry (SOKA) model and the Brunswik Lens Model, which explain how observable cues differ in visibility and diagnostic value. She highlighted the distinction between deliberate identity claims and unconscious “behavioural residue,” noting that digital environments, much like physical spaces such as bedrooms or offices, generate extensive residue that can be analysed at scale.

Jo presented research showing that, since 2015, algorithms have outperformed humans in assessing digital profiles. While this capability offers powerful insights, she emphasised the ethical risks it poses, including bias, lack of consent, opacity, and misuse. The Cambridge Analytica scandal served as a cautionary example of profiling used to influence behaviour at scale. Jo concluded by stressing that ethical digital profiling requires triangulation across data sources, cultural sensitivity, continuous model retraining, and an understanding that algorithms produce predictions, not definitive truths.

The Cambridge Analytica scandal served as a cautionary example of profiling used to influence behaviour at scale.

The session, led by Dr Heather Shaw, shifted focus to behavioural profiling and biometrics. Heather demonstrated how seemingly mundane digital behaviours—such as unlocking a phone or switching between apps—can reveal consistent behavioural signatures over time. Grounded in interactionist psychology, her work shows that behaviour is shaped by both individual traits and situational context. People behave differently across digital environments, yet those differences are themselves stable. Deviations from established patterns may signal meaningful change, but Heather cautioned against simplistic interpretations, as context often explains variation. Participants took part in a hands-on machine learning exercise, modelling behavioural data to explore how individual signatures can be identified.

Then, Heather introduced a digital profiling card game to help participants discursively explore how different combinations of data, including, demographic, employment, culture, and health data could be used. These combinations were examined in relation to how they support or enable targeted advertising, discrimination, surveillance state practices, and massively invasive privacy breaches. The session highlighted both the sophistication of current profiling techniques and the importance of theory-driven, ethically informed analysis to avoid harmful false positives.

Day two opened with Professor Oli Buckley, who brought a security intelligence perspective to digital profiling. Oli outlined the economic value of personal data, illustrating how stolen digital assets—from streaming accounts to complete digital identities—are traded in illicit markets. He introduced the concept of “digital shadows”: the cumulative data traces individuals leave behind through everyday interactions. Platforms can build detailed profiles in remarkably short periods, often using micro-engagements that users barely notice. Oli explored behavioural biometrics such as keystroke dynamics, showing how timing patterns can be used for identification and authentication. Although these techniques have applications in fraud prevention and insider threat detection, Oli emphasised the ethical risks associated with predictive security systems, particularly bias, explainability, and the real-world harm caused by false positives.

Oli’s social deduction game, Cold Reads and Hot Takes, allowed participants to explore insider threat detection through role play. Participants’ gameplay highlighted how narratives, bias, and incomplete information shape perceptions of risk, reinforcing the human dimension of profiling.

The closing panel discussion and audience Q&A addressed broader questions about acceptability and responsibility. How authentic do synthetic datasets need to be? How many signals are enough to make profiling robust? Is more data always worth the marginal gain? Across discussions, there was consensus that profiling must be theoretically grounded, explainable, and ethically justified. Digital profiling offers significant potential, from improving cybersecurity to advancing psychological research, but it also poses profound ethical challenges. As repeatedly emphasised throughout the event, algorithms predict—they do not conclude.

For early career researchers, the event underscored the importance of interdisciplinary collaboration, combining psychological theory, computational methods, and ethical scrutiny. As digital data and AI capabilities continue to expand, the critical question is not only what digital profiling can do, but what it should do.