Data
Understanding key challenges in data and AI: insights from the 'Data Peer-to-Peer’ roundtable
02 October 2024 • 3 min read
AI is everywhere. Good AI – AI that feels genuinely intelligent, useful and valuable – is a little less commonplace. Recently, we hosted a ‘Data Peer-to-Peer’ roundtable in London, bringing together leaders from various sectors, including public sector, financial services, and food & beverage. The group came together to share their perspectives on the current challenges facing organisations wanting to up their Data & AI game.
These were the major themes from the event.
1. Connecting data quality and AI effectiveness
We’ve been talking about ‘garbage in, garbage out’ in a tech sense since the 1960s, but AI shines a harsh new light on the old idiom. Bad, inaccurate, incomplete or biased data leads to flawed (and sometimes downright discriminatory) AI. Compounding this problem is that the outcomes of AI – the generative solutions it delivers – are frequently highly visible. Flawed AI isn’t just a data problem; it’s often a reputational one too.
High quality data requires robust data governance frameworks that eliminate not only error but potential bias. And a robust data framework needs clear objectives, expert quality management, categorised data, and clear data ownership (among other elements), all tied to a data-driven culture that drives the right ambitions, behaviours and outcomes.
2. The role of data and AI in understanding consumer needs
Customer journey mapping. Behaviour analysis. Predictive analytics. Data and AI can have a powerful role to play in identifying patterns and optimising processes. One specific takeaway from the roundtable was the value of cross-sector collaboration as a way of gaining fresh insights into consumer behaviour and preferences, and enhancing personalised experiences in B2B and B2C contexts.
3. Adopting ethical data and AI
We’ve already explored the importance of data quality in supporting ethical AI. Regulatory compliance and maintaining trust should also be prioritised while exploring AI's potential.
The roundtable also considered the value of the initial commitment to apply ethical AI, ensuring it aligns not just with organisational values but societal norms. Whilst important for every sector, ethics was seen as a particular priority in ‘sensitive’ industries such as banking and healthcare.
Appointing ethics champions and maintaining human oversight of AI were also considered important to boosting bias reduction and responsible decision-making. They are also vital to maintaining trust by giving customers reassuring transparency in AI practice.
4. Achieving ROI on data and AI investments
Inevitably, the roundtable discussed the ROI of AI and the importance of aligning AI initiatives with robust, industry/company-specific business goals.
In banking, for example, AI-focused metrics may focus on customer satisfaction rates and successfully completed chats with AI-powered chatbots. It may target the number (or value) of successfully detected fraudulent activities or the accuracy of credit risk assessments.
Understanding the ROI metrics from the outset of an AI initiative can help move AI projects from the experimental phase to scalable, production-level solutions, especially in complex, regulated sectors.
5. How to balance external partnerships with in-house data/AI fluency development
The issue of in-house vs external is one extremely familiar to AND Digital. It’s something that has informed AND’s way of working for a decade, and AI is no exception. The roundtable explored the tension between developing in-house expertise and involving third party experts who may offer fresh insights on AI but lack the deep understanding of an organisation’s values.
The roundtable discussed the value of a balanced approach, leveraging in-house expertise and external partnership support to achieve real innovation and customisation in AI integration. Third party engagement should run parallel with comprehensive internal training programmes focused on AI and data fluency, empowering employees across various roles to adapt to technological changes and enhance their skill sets.
Flexible partnerships with AI subject matter experts should place ethics at their core, with compliance with internal policies and ethical standards a central part of the relationship.
Our ‘Data Peer-to-Peer’ roundtables are intimate events that bring together senior leaders from multiple industries to network and discuss core topics around data. If you’re interested in joining us at a future event, please contact us here.