Tech Talk Interviews
From Compliance to Culture: Rebecca Gallagher on Responsible AI
An Interview with Rebecca Gallagher, AI Compliance Manager at Weir Group PLC
At The AI Summit London, we spoke with Rebecca Gallagher, AI Compliance Manager at Weir Group PLC, who shared her perspective on how to move beyond tick‑box compliance to embed ethics in day‑to‑day decision‑making. She urges tailoring frameworks to culture and risk appetite, building multi‑lens talent, and resisting short‑termism. She also highlights human‑centric governance, enterprise‑wide accountability and clear “nutrition labels” to support transparent, context‑aware AI."
Read the Full Interview
Interviewer: How can organisations move beyond compliance to embed their ethics and values?
Rebecca: Compliance is something we have to do; ethics is something we should do. My first recommendation is to avoid copying and pasting frameworks and principles from other companies. It is crucial to define what works for your organisation’s culture and AI strategy. Within that strategy, be clear about your AI risk appetite. While there will be common ground with peers, you must tailor your approach to your specific needs.
Another key point in this technological revolution is building multi‑lens talent. For example, data scientists developing AI models may have a strong technical lens, but if their work affects people, they also need a strong ethical lens to understand potential impacts and how to mitigate associated risks. Finally, lead with a sense of duty: prioritise doing the right thing over short‑term gains, which can be tempting in a fast‑moving landscape.
Interviewer: What does implementing or incentivising long‑term gains look like, and how do you ingrain that mindset?
Rebecca: Short‑term cost saving can be a business illusion. If customer and partner confidence is sacrificed for short‑term gains, companies can find themselves in risky situations. A key topic is avoiding the talent trap. It might look attractive on quarterly balance sheets to reduce headcount as you embed automation, but if you extract human talent and the organisation’s historical knowledge, you bleed the company of the core expertise needed to scale AI safely and sustainably over time.
There is also a lot of unproven technology in the market. We have to ensure what we roll out is fit for purpose and proven in our context. A solution may work elsewhere but not necessarily in our organisation. When evaluating short‑term gains, ask whether they align with your AI strategy and long‑term needs.
Interviewer: How important is it that a company first defines its values and what it stands for?
Rebecca: Most organisations have core business values, long predating AI. Those are your North Star. But you need to double‑click on them to understand what they mean in practice in the AI landscape—for customers, culture and employees—as you roll out AI. It requires a deeper look at business values with an added AI risk lens.
Interviewer: How can companies demonstrate integrity towards customers and employees?
Rebecca: In a fast‑changing landscape, companies must stay agile while maintaining strong governance foundations. We focus on a human‑centric framework with clear accountability across the business so AI risk is owned enterprise‑wide, not buried in technical or compliance teams. That ownership is critical.
You also need robust structures and forums that foster a culture of collaboration. At the same time, ensure the right people are at the right table at the right time to ask the right questions, provide guidance, identify risks and surface opportunities. To make this work, invest in strong AI literacy programmes so people at every level are empowered to make sound decisions.
Interviewer: What are some of the right questions that should be asked?
Rebecca: At a strategic level, when a business area wants to procure an AI solution, ask whether the use case aligns with the company’s AI strategy. What business value will it deliver, and what risks does it introduce? Is it within our risk appetite? Are risks properly understood and proportionately controlled? Is there appropriate human oversight of the use cases being evaluated? To answer these accurately, you need the right level of knowledge and expertise.
Interviewer: What steps can organisations take to improve transparency and readiness?
Rebecca: As AI is democratised and more people interact with it, we must simplify. Translate complex data science into human‑readable information that everyone can understand. For example, use simplified “nutrition labels” to explain a model’s risks, potential biases, likelihood of drift and limitations. This is vital for non‑technical buyers—say, someone in HR procuring a third‑party AI tool—who need clarity on limitations and bias to determine the necessary level of human oversight of outputs.
Context is also king. Make sure the AI solution you deploy is fit for purpose in your setting. A system tested in a factory in Detroit may not be appropriate—or may pose additional risks—for a factory in London. Understand your context, purpose and the limitations of each solution.
Interviewer: AI is evolving quickly and the key trends seem to change weekly. What do you see as the key trends now?
Rebecca: We have moved from the “wow” phase to the “how” phase—from AI assistants to automated agents. With automated agents, we need strong governance and oversight. We must embed the people, processes and tools—and a culture of collaboration—to ensure safe, sustainable rollout. Returning to AI literacy, employees and teams need to be empowered with the right information to make sound decisions. That is how you scale AI, deliver business value and manage it safely and sustainably over the long term.















