“Being at the forefront of technology has been a key thing in this project for us.”
— Ashfaq Farooqui, RISE Sweden
Ashfaq Farooqui from RISE Sweden speaks about the unique role his organisation plays in the HIVEMIND project as leaders of the AI-enabled Design-by-Contract Programming work package. He reflects on what makes the project stand out, the technical and societal challenges involved, and how tools developed in HIVEMIND could shape the future of reliable, safe AI systems.
Watch the interview
Prefer reading?
You can find a full transcript of the interview below.
My name is Ashfaq Farooqui, and I’m from RISE Sweden. We are an independently owned organisation with a mandate to bridge between industry and academia — across both private and public sectors.
We are particularly focused on the verification and validation of systems, and we have a strong interest in cybersecurity and safety, especially the interplay between them. That’s essentially our area of work.
We have people working on standards, and others developing methods to evaluate those standards, both for evaluating technologies and for guiding their development. In the HIVEMIND project, we are involved in Work Package 6, where we will be developing tools for the different large language model (LLM) agents that will be tested in the use cases.
What excites you most about being part of the project?
There are multiple exciting aspects. One is the consortium itself, all the partners are strong in their respective areas, and learning from one another has been a major driving force in the project.
And of course, the technology. We’re living through this massive wave of interest, a kind of bubble, and the hype around it is huge. Everyone wants a piece of the cake. For us, being at the forefront of technology has been a key part of this project.
What impact do you expect HIVEMIND to have on industry and society?
In terms of impact, we believe this project can directly contribute to the development of new technologies and products that are correct by design, meaning they work as intended from the start, without requiring extensive testing just to ensure they are safe.
Through HIVEMIND, we aim to develop tools that lead to more reliable systems, better in both design and implementation. This is essential for a future where applications and technologies emerge so quickly that standards and regulatory bodies can’t keep up.
Having systems, and AI tools like LLMs, that support trustworthy development from the beginning is something I think will have a significant societal impact, if approached properly.
What challenges lie ahead for the HIVEMIND project?
Technically, there are many challenges. I believe that with enough time and resources, most of them can be solved. However, there’s also a broader challenge around the use of AI itself. We invest massive resources into building AI systems, and we have to ask ourselves whether we’re doing enough to reduce the environmental impact of that.
If we don’t think carefully about how we build these systems, we risk undermining the very innovation we’re pursuing. We’ve seen that happen before, in the AI winters of the 1990s and early 2000s. So we need to keep reflecting on the ethics and background of AI systems. That said, I’m confident that we’ll move forward and solve these problems over time.