Overview
What happens when AI goals and human values diverge? Explore AI alignment, online safety, and trust in the digital age.
About the Talk
AI alignment is often described as the challenge of ensuring that AI systems behave in ways that are consistent with human intentions and values. Many of today's online safety concerns can be understood through the lens of this challenge. Whether it is a chatbot generating misleading information, a recommender system amplifying polarising content, or an AI agent pursuing goals in unintended ways, the underlying challenge is often the same: a gap between what an intelligent system is designed to do and what people expect and value.
This talk explores AI alignment through the lens of online safety and from the perspective of the SPRITE+ concepts of Trust, Identity, Privacy, Security, and Safety (TIPSS). It examines why online safety is increasingly being viewed as an AI alignment challenge, revealing the important relationship between AI systems, human values, and societal expectations.
Ultimately, the talk considers a simple but important question: if people do not always agree on the values that shape society, how do we ensure that increasingly powerful AI systems act in ways that reflect them?
About the Speaker
Dr Edward Apeh is a researcher, educator, and SPRITE+ Expert Fellow. He is also a Principal Academic in Computing at Bournemouth University. His work explores how artificial intelligence (AI) is reshaping trust, safety, identity, and human decision-making. His research spans AI safety, cybersecurity, online harms, digital identity, fraud prevention, and the design of trustworthy digital systems.
Edward works with academics, industry, policymakers, and communities to better understand how people interact with AI in the real world and how technology can be designed to support, rather than undermine, human wellbeing. Through his work with SPRITE+, the Financial Conduct Authority, and a range of other UK and international research initiatives, he has become increasingly interested in the societal challenges of building trustworthy and responsible AI systems.
His current work focuses on a critical question emerging from society's growing engagement with AI:
What does it really mean to align AI with human values in a world where trust, identity, privacy, security, and safety are constantly evolving?
