By Judith Hawkins, 13 October 2025
ICA’s inaugural AI Week saw hundreds of delegates unite to explore how AI is transforming the compliance and financial crime landscape.
Across the three-day virtual event, more than 30 experts shared their insights and advice on AI best practice within compliance, risk management, and financial crime prevention.
Speakers from around the world provided guidance on how to keep pace with the latest AI innovations, while also maintaining effective oversight and alignment with regulatory expectations.
If you weren’t able to attend ICA’s AI Week, you can still purchase a ticket to watch all of the sessions on-demand, with recordings available until 18 December.
Key themes
An overarching theme seen across the mix of panel discussions and practical case studies was the value of taking a collaborative approach to AI implementation, by enabling ongoing dialogue between compliance and technology teams. A pivotal factor to achieving this was ensuring cross-departmental AI literacy, with an ability to ‘speak the same language’.
Another core message reiterated by many of the experts was to always identify business needs first, before then looking into AI solutions that can specifically address these identified requirements (as opposed to just adopting AI for the sake of ‘keeping up’).
Taking the time to build up trust in chosen AI tools amongst both internal and external stakeholders, particularly via establishing strong AI governance, was another key takeaway of the event.
Regulation, governance and policy
The first day focused on AI regulation, governance and policy, starting with a panel discussion considering what good governance looks like, chaired by ICA Vice President Tim Tyler.
The panellists highlighted how compliance professionals’ existing skills mean they are already well-placed to support their firms with AI. Dean Cooper, Senior Compliance Manager at Unum, said: ‘As the technology moves so quickly, I think it’s really important to lean on one of the bread-and-butter skills in compliance, and that’s horizon scanning.’
Following sessions on the first day considered the importance of ensuring explainability of AI tools. In an informative presentation on how to provide meaningful explanation for AI decisions, Lucie Novotna, AML Solution Engineer at Resistant AI, demonstrated how highly technical AI models could be broken down and identified as recognisable transaction monitoring activity.
A best practice case study in setting up AI risk committees and centralised oversight structures was then provided by Lindsey Matthews, Chief Risk Officer at the Universities Superannuation Scheme. He talked through the stages of their AI governance journey, including establishing clear guardrails, creating an AI steering committee comprised of key stakeholders, and embedding AI policy into existing oversight frameworks.
Financial crime focus
AI in relation to financial crime was the focus of the second day, as thought leaders examined how the technology is being used by both good and bad actors in this space, and what firms must do to stay ahead of the criminals.
ICA President Pekka Dare started the day talking to Neil Jennings, Founder of GLF Strategic Compliance, on ways to determine whether AI tools adopted in financial crime compliance are fit for purpose.
Jennings emphasised that first and foremost businesses should focus not on the AI tools themselves, but on ensuring that you ‘know your needs’ (KYN). He explained that this was a four-step process: knowing your landscape, your processes, your current tools, and the gaps. Only then can firms ensure they are selecting AI tools that effectively address these gaps.
As one example of how business needs were prioritised, CCO & MLRO Jason McKinney shared his experience at My EU Pay of building and implementing a bespoke AI model for transaction monitoring, rather than relying on ‘off-the-shelf’ solutions. He also covered the importance of both upskilling and hiring the right people to create a strong mix of both compliance and technical skills, and the benefits of setting up a shared model risk team.
Collaboration continued to be a theme as the AI event took time to consider where cybersecurity and compliance risk intersect, and how CISO-MLRO alignment can help manage joint threats from cybercriminals. As noted by Ellen Watson-Hicks, Chief Risk & Compliance Officer at Railpen: ‘I think the risks of not collaborating are just far too great at the moment.’
Throughout the day experts also discussed the ongoing ‘AI arms race’ between legitimate businesses and criminals. In a panel considering how to stay ahead of criminal innovation, AI detection tools to spot adversarial use of the technology were highlighted as a must in the current financial crime landscape.
Data, technology and innovation
On the final day of the event, attention turned to how cutting-edge AI technologies are being applied to compliance challenges, and the critical role of data quality in ensuring effective, responsible deployment.
ICA’s Tim Tyler kicked off proceedings again, with a discussion on the factors driving innovation in AI, alongside Adrianna Fabijanska, Global Head Financial Crime Compliance, Investment Banking at ING, and John Clarke, UK MLRO & Global Head of Financial Crime Compliance at Schroders.
Driving factors identified included the current over-reliance on manual processes, high false positive rates, opportunities for greater efficiency, scalability and precision, and rising regulatory expectations around use of technology.
A helpful case study was also presented by Onyinye Asala, International Projects Compliance Director at Ciena, who explained the practical steps she had taken to co-design agentic AI solutions in collaboration with data teams, including resolving any data gaps.
The opportunities which come with AI innovation were further covered when another of the AI Week’s moderators, Tony Tarquini, Founder & CEO of 5189 Limited, spoke to James Booth, Head of AML, Counter Terrorism and Sanctions at Silent Eight.
Booth described how AI is enabling previously reactive compliance teams to now proactively address alerts in real-time, moving from ‘firefight to foresight’.
The potential for AI and its effective governance to be a business enabler was then outlined by Jimmy Lin, Chief Product Officer at SAI360, who explained why firms should be focusing on building stakeholder trust in AI, both internally across their business and externally with customers and regulators.
The importance of securing stakeholder trust was underlined in the final panel session, with Maria Lukashova, Compliance Lead at Microsoft, highlighting the value in widespread AI training and support so that non-technical staff can also build their understanding and confidence in AI tools. By considering lessons learned when deploying AI in compliance, this concluding discussion wrapped up an insightful AI Week with more examples of best practice to take forward.
Find all of these virtual sessions and more by registering for on-demand access to ICA’s AI Week now.