Leah Claireaux (Senior Cyber Research Specialist, BT)

Agentic AI for Cyber Security

AI has advanced drastically over the last few years with the development of GenAI. This talk will cover some of the latest cutting edge cybersecurity research in BT, in particular how we are looking to leverage these advancements for Agentic AI applications with Large Language Models (LLMs).

Leah Claireaux, over 10 years experience in telecommunications industry & Security Professional of the Year at UK IT Industry Awards 2023. Background expertise in AI, cybersecurity, network and software engineering. Currently leading a team to deploy cutting-edge artificial intelligence algorithms, to protect BT's network.

Al Brown (Associate Fellow, Royal United Services Institute)

Warbots - a View from the Trenches

“Artificial intelligence is the future, not only for Russia, but for all humankind.” .. “Whoever becomes the leader in this sphere will become the ruler of the world.” Vladimir Putin

Robotics and artificial intelligence are already being employed in conflict. However, artificial intelligence manages to sit at the peak of ‘inflated expectations’ on Gartner’s technology hype curve whilst simultaneously being underestimated in other assessments. Artificial intelligence and robotics are technologies without a monopoly. Almost anyone can take emerging advances in these technologies and apply them to conflict. And they are. But which adaptations are successfully exploited is governed by hard scientific facts and the gritty realities of employment in the field.

So, what are the likely effects on conflict of the trends in artificial intelligence, robotics, economics, data and society? And what do people commonly get wrong - often with total certainty? This talk will take you through what is and is not likely to change in warfare, and what has already changed in conflict.

Al Brown is an Associate Fellow at the Royal United Services Institute. Al is a former Research Fellow at Oxford University and Chief of the General Staff Scholar researching perception, information integration, and optimisation for competitive advantage in biological, digital, and organisational systems. Al was previously the lead for UK Defence on the study of global strategic trends in robotics and artificial intelligence and its impacts on conflict. He authored a short book on this and has been one of the Group of Government Experts providing advice to and speaking at the United Nations. He writes and has often lectured on related subjects at the Alan Turing Institute, the Royal United Services Institute, Oxford University, UCL, and Cranfield University. Al is also a former Army Officer. He commanded a bomb disposal and high-risk search Regiment, led the Counter-IED Squadron in Helmand, and has deployed on operational tours in Afghanistan and the Balkans.

Prof Niels Lohse (Professor of Manufacturing Automation and Robotics, University of Birmingham)

Co-AIMS – AI for People-Centric, Regenerative, and Resilient Manufacturing

Industry 5.0 represents a shift from pure automation toward a more human-centred, sustainable, and resilient manufacturing paradigm. This presentation explores how Artificial Intelligence can be leveraged to create people-centric, regenerative, and resilient manufacturing systems that work with—and for—humans. We’ll start with an overview of the Industry 5.0 vision, which prioritises empowering workers, reducing environmental impact, and enhancing system resilience against disruption.

We will discuss the real industrial challenges faced on this journey: the cost and complexity of integrating AI into legacy systems; the need for trustworthy, explainable AI to foster human-AI collaboration; and the barriers around skills, data sharing, and verification. Through selected industry use cases, we’ll demonstrate how AI can enable better synergy between people and machines—from adaptive skill learning for in-process optimisation to ramp-up optimisation.

Finally, the talk will introduce Co-AIMS, the new £11 million EPSRC Manufacturing Research Hub launching this October. With over 50 industrial partners, Co-AIMS will drive the UK’s leadership in collaborative AI for manufacturing sustainability. Its mission: to deliver robust, safe, and trusted AI systems that boost productivity while eliminating waste, reducing emissions, and empowering human workers. By framing manufacturing as a socio-technical ecosystem of people and AI, Co-AIMS aims to realise the promise of Industry 5.0—not just automation, but augmentation that delivers sustainable value for industry, society, and the environment.

Professor Niels Lohse is Professor of Manufacturing Automation and Director of the Co-AIMS Manufacturing Research Hub at the University of Birmingham. He specialises in the design and integration of advanced manufacturing systems, with over two decades of experience bridging industry needs and cutting-edge research in manufacturing automation, robotics, systems engineering, and AI-enabled production.

Professor Lohse’s work focuses on making manufacturing systems more agile, sustainable, and human-centric. He leads interdisciplinary teams developing methods for digital twins, smart machine integration, and collaborative human-AI systems that support resilient, flexible production. He has a strong track record of working with industry to translate research into practice, tackling real-world challenges of interoperability, adaptability, and sustainability in advanced manufacturing.

Dr Cailean Gallagher (Organisation Co-ordinator and Associate Lecturer in Management, University of St Andrews/Workers' Observatory)

How are Gig Workers Resisting AI-driven Control?

Platform-mediated gig work such as riding for Deliveroo or driving for Uber is widely known to be exploitative. Conditions have kept deteriorating with the introduction of AI-managed fares (dynamic pricing) and increasingly sophisticated surveillance. This includes live GPS tracking of riders to calculate delays, detours, AI-deduced suspicion of combined work, and AI-expedited decisions to impose sanctions. On the back of these systems of surveillance and control, food delivery apps progressively aggregate incredibly granular data about the cities they operate in and how riders navigate them. Once the apps reach a level of market dominance, they turn this data – both the urban grid and the live-tracking – into intensely disciplining AI tools. The data is never disclosed to workers or researchers. The decisions can’t be appealed. Workers’ efforts to understand how platforms drive down pay and conditions are hampered by companies’ determination to conceal data on workers’ movement and payments. What can enable gig workers to counteract data imbalances in their struggle to resist exploitation? We will talk about experiments conducted in Edinburgh, involving a workers’ inquiry with South Asian and Spanish-speaking food delivery riders, providing workers with tools and training to create and distribute a survey researching their conditions. Through the process of its design and distribution, some of the impact of this data and associated disciplining came into sharp relief. In the process of building the survey and the discussion around working conditions and tracking that it facilitated, a collective started forming. We draw on this and other experiments to share the concept and discuss the merit of participatory workers’ data science (Gallagher et al, 2024), as a method not only of building research, data and awareness, but of developing the workers’ inquiry tradition as a creative means to foster collective organising and bolster workers’ power where they are deeply disenfranchised.

Gallagher, C., Gregory, K. & Karabaliev, B. (2023) Digital worker inquiry and the critical potential of participatory worker data science for on-demand platform workers. New Technology, Work and Employment, 1–21. https://doi.org/10.1111/ntwe.12286

Dr Cailean Gallagher is the co-ordinator of the Workers’ Observatory. He is an Associate Lecturer at the University of St Andrews, researching and teaching about changing forms of labour control and resistance. He is also the editor of the Scottish Left Review.

Michael Free (AI Research Manager, BT Digital)

LLMs in Industry: Success Stories and Challenges

The talk will include practical lessons learned from deploying generative AI for the first time, what works, what didn't and what are LLMs best used for? The talk will also cover some cautionary tales and fundamental challenges that mean LLMs aren't great for everything!

Michael Free has been an AI researcher with BT for almost a decade, working on applications of everything from Convolutional Neural Networks to Large Language Models. His role involves forming a bridge cutting edge research from academia and business applications by working on proof of concept applications and furthering the state of the art in specific application areas. His current interests are around LLM interpretability for application safety and task specific fine tuning of LLMs.