Human Traffic and the limits of AI
ABSTRACT: ‘Insider Threat’ is a formidable risk to business because it threatens both customer and employee trust. Accidental or malicious misuse of a firm’s most sensitive and valuable data can result in customer identity theft, financial fraud, intellectual property theft, or damage to infrastructure. Because insiders have privileged access to data in order to do their jobs, it’s usually quite difficult for security professionals to detect suspicious activity; a process even more challenging to automate (and deploy at scale across the large organisations that most need it) as – so I will suggest – computers fundamentally lack semantic understanding of the meaning of the ‘bits’ they so adroitly process. Conversely, in this talk, I will outline a new approach to ‘Insider Threat’ detection that draws inspiration from the `Traffic Analysis’ of encrypted `Axis signal traffic’ undertaken at Bletchley Park in WW2. A novel approach that (i) conceives companies as complex autonomous autopoietic entities and (ii) deploys state of art computational analysis of the communication flows that so define the company to flag potentially aberrant employee behaviour; intelligence that can be leveraged to help detect HR problematics before they arise.
BRIEF BIO: J. Mark Bishop is Director of The Centre for Intelligent Data Analytics and Professor of Cognitive Computing at Goldsmiths, University of London and a co-opted expert on AI for ICRAC (the International Committee for Robot Arms Control). In these roles Mark has been awarded research funds in excess of £3m and published around 200 articles in the field of Cognitive Computing, addressing (a) theory – where his interests focus on the foundations of the swarm intelligence paradigm “Stochastic Diffusion Processes”; (b) application – where Mark has led numerous real-world projects (ranging from Procurement & HR Analytics and Fraud detection to Autonomous Robotics, Advanced Neural Networks and Space Exploration); and (c) its philosophical foundations – where he has explored the limitations of AI and Neural Computation, most famously developing the “Dancing with Pixies” reductio ad absurdum (a powerful and original argument against the possibility of machine consciousness). In addition, Mark has edited three collections of essays: together with John Preston, a critique of John Searle’s famous argument against machine intelligence, “Views into the Chinese Room” (OUP, 2002); with Andrew Martin he co-edited a collection of essays on “Contemporary sensorimotor theory” (Springer 2014) and with Experience Bryon et al., Mark co-edited a volume on “Embodied Cognition, acting and performance” (Routledge 2018).