Building an Early-Warning System for AI Dysfunction with Dr Homa Molavi

Remember meeting Dr Homa Molavi in our recent post? Today we hear more about the research she is currently undertaking. Focusing on design thinking and systems thinking, Dr Molavi and her research lab are examining how organisation and governments can approach AI integration in a structured, safe, and fit-for-purpose way rather then purely trend driven or reactively. Dr Molavi explains the research question that provides the ‘golden thread’ for her explorations:

At its heart: how can we use AI and data analytics to make organisations, institutions, and communities more sustainable, resilient, and accountable? Whether I’m looking at a farmer’s decision about water use or a corporation’s response to a reputational crisis, the underlying question is always about behaviour, governance, and the conditions that enable better decisions.

Dr Molavi goes on to share that one of the big interests for her is how organisations represent, or rather, misrepresnt their AI use- that is, how AI actually surfaces hidden behavioural patterns. You know how some companies have ‘green-washed’ – made themselves appear more sustainable than they are to look more attractive to consumers and employees? Dr Molavi recognises that the same happens with AI – given all the claims about how we are in the ‘age of AI’, organisations may feel pressured to put forth the perception that they are not being left behind. Dr Molavi reveals the significance of the problem:

The speed at which AI tools are being adopted far outpaces the frameworks designed to govern them safely, and the gap between what organisations claim about their AI and what is actually functioning underneath is often startling. Those surprises have, if anything, made the research more urgent.

Of course, our Dr Molavi is not just describing this phenomena, she’s creating structured approaches to help:

We’re developing diagnostic frameworks that can identify AI-related risks early: governance failures, inflated capability claims, technical debt accumulating silently beneath the surface. Think of it as an early-warning system for AI dysfunction in organisations, before it becomes a crisis.

This isn’t all she’s working on. She goes on to say, “I’m also developing work that bridges the gap between academic research and SMEs smaller organisations that want to adopt AI responsibly but lack the in-house expertise to do so safely. Knowledge transfer in that space feels like one of the highest-impact things I can do.” This work is important. As Dr Molavi points out:

Because the challenges that cities, businesses, and communities face, whether it’s water scarcity, climate adaptation, institutional fragility, or AI governance, are not purely technical problems. They are behavioural and governance problems. Getting the data analysis right, and connecting it meaningfully to policy, can make the difference between interventions that genuinely work and ones that simply look good on paper.

Our polymath uses a range of methodological approaches in her work, gained from years of teaching research methods and supervising hundreds of student dissertations. Each of these student research projects, with different research questions, require different approaches and solutions – this experience, this breadth, has shaped and honed Dr Molavi’s capabilities. As Dr Molavi reflects, “That said, I’m drawn most strongly to quantitative methods: working with large datasets, NLP, machine learning, and AI-driven analytical tools. But I’m also a genuine believer in interview-based methods. Numbers tell you what is happening; conversations tell you why. Both matter.”

She is currently working on a number of publications, including journal submissions and practical frameworks for direct use by organisations and policymakers. Dr Molavi spreads her workthrough conference presentations and chairing, collaborative research with applied institutions, and direct consulting and advisory work with organisations. She has recently chaired the A.H.E.A.D. Summit roundtable at the Univesity of Manchester, which brought together academics, industry innovators, and investors to think about how early-career talent can bridge the gap between research and real-world impact; “That kind of convening getting the right people in the same room  is something I care deeply about”. Dr Molavi emphasises that her work is not just an academic exercise but one that will influence policy, practice and public understanding:

By demonstrating that AI is not just a tool for efficiency, it’s a tool for finding emerging issues, surfacing hidden patterns, and supporting sustainability, if used thoughtfully. I want practitioners and policymakers to stop seeing computational methods as a black box, and start seeing them as a way of making decisions more evidence-based, equitable, and honest. And I want businesses and governments to adopt AI responsibly with a clear-eyed understanding of the reputational and operational risks involved, not just the opportunities.

I want policymakers to have better tools, and I want those tools to come from evidence rather than hype. Research that sits only in academic journals has limited value. The bridge between findings and actionable policy is where I want to spend my energy.

Alongside several articles in press, Dr Molavi is also planning engagement events and workshops aimed at practitioners and policymakers, particularly around responsible AI adoption for smaller organisations. Keep an eye out on NUSC communication channels for more details!

Ultimately, the aim of Dr Molavi’s research is to shift how we think about AI – that it is not just a technical or financial decision. As she adds, “the evidence suggests that organisational culture, governance structures, behavioural patterns, and leadership decisions matter far more than the technology itself in determining whether AI is adopted responsibly and effectively. That reframing has significant consequences for how we regulate AI, how we train leaders, and how we design the organisations of the future.”

The cover image on this post was generated using WordPress built in AI function. No AI was used in the writing or reviewing of this blog post. 

Discover more from Networks and Urban Systems Centre (NUSC)

Subscribe to get the latest posts sent to your email.

Leave a Reply

Your email address will not be published. Required fields are marked *