Keynote: Data Transparency, AI Use Cases, Data Sovereignty
by Dr. Phil Winder , CEO
At Winder.AI, we’re seeing a shift in how businesses are adopting AI—not just for innovation, but for real, tangible commercial outcomes. I had the privilege of sharing these insights as a keynote speaker at ITAPA in beautiful Bratislava, Slovakia. The audience was looking for an insight into how AI is being used and some of the key challenges that are being faced today. I took the opportunity to share some of my thoughts about the importance of data transparency, some interesting use cases, and future regulation to watch out for.
Below is a recording of the talk, with a written version below.
AI in 2024
While the media loves a sensational “failure rate” headline, the reality is that when businesses align AI projects with specific goals, the results are transformative. From cost savings to new revenue streams, AI is creating value like never before. More than 70% of business say that they are using AI.
Data Cleanliness and Management
Even today, in the era of GenAI, data is still the most important thing. It’s the foundation of all our projects. The cleanliness, which in general means the quality of the data, dictates how predictive trained models are and how much they can be trusted. Even in generative AI, unclean contexts can lead to LLM inaccuracies and hallucinations.
Earlier in the day there was a demo from RedHat showing an example image classification model. But I was struck by how much of the demo was taken up moving and organizing data. I decided to highlight that in my talk as evidence that points towards how much effort is required to manage data. When considering regulation, data management is even more important as a key component in the compliance process.
Start small. Focus on one well-defined dataset, improve its quality, and measure its impact. This process builds trust in AI outcomes over time.
Data Transparency
Trust is the common foundation of all relationships. Democracy, and the relationship between the people and their government, is based on trust.
One of the best ways to build trust is through transparency. Governments and businesses alike are increasingly using data transparency to build relationships.
Does the North-South Divide Still Exist in the UK?
To demonstrate how data can shift perspectives, I used a dataset from the UK government showing the average income of people in different regions. The data is interesting because it is very fine-grained geographically. I took the data from the ONS and used it to create a map of the UK showing the average income of people in different regions. You can interact with this map online.
In this map you can see that there is an income disparity between greater London and the south east of England compared to the rest of the country. But this is not the whole story.
I then took another dataset from the ONS that represented the average housing costs in each of these regions. I then subtracted housing costs from income to get a measure of disposable income. You can interact with this map here.
This map shows a slightly different story. The red band of high income has now grown to include much more of the south of England and interestingly reaches up towards Birmingham and even has far as the south of Manchester. Some areas in Yorkshire are also on par.
Seen in this light, the greatest disparity is now between Wales and patches of the east coast of England and everywhere else. This is a very different story to the first map and casts doubts on the traditional narrative that the north-south divide is still a thing.
This isn’t just an academic exercise. Businesses can leverage similar analyses to uncover hidden trends in their customer data, optimize regional marketing strategies, or identify new opportunities for growth.
Corruption Index
The previous demo is a quick example of “citizen data science.” Charities and NGOs are increasingly leveraging public data to produce derivative data like Transparency International’s Corruption Perception Index. This index transforms public datasets into a measure of how corrupt different countries are perceived to be.
Since they themselves publicise their data, I again can analyse this data to answer my own questions.
Corrupt UK’s Rise and Fall
I was interested to learn how recent events affected the world’s perception of the UK, since I’ve heard murmurs of a few people I respect talking about institutions like the BBC falling from grace.
So I took Transparency International’s data and plotted the UK’s data against the US and Ireland.
Alongside some speculative events, you can see that the USA took a hit around when Trump became president for the first time and the UK took a hit just after COVID. Ireland on the other hand has seen it’s rank jump considerably over the past few years.
Transparency Feeds Transparency
The previous example shows how transparency can compound and snowball. New analyses bring fresh ideas. And some of those ideas might literally change the world.
We often talk about innovation in business, but this is represents innovation in the public sector. The next big policy idea might come from a student project or a charity in another country. There’s so much value in this data it would be silly for governments not to release it.
How to Make AI Successful
- Focus on business impact: Align AI projects with measurable goals (e.g., cost reduction, revenue growth).
- Leverage scalability: Use AI to tackle problems that humans alone can’t scale efficiently.
- Data drives outcomes: Invest in clean, representative datasets to ensure reliable, trusted AI models.
- Enhance human capabilities: Design AI to work alongside employees for better decision-making.
- Adopt ethical AI: Build trust with customers, employees, and regulators through fairness and transparency.
AI Regulation
The EU AI Act is a really important piece of regulation. Not because of what it is, but because of the direction that it is signalling. The regulation of the exposure of AI solutions is now here and companies need to take it seriously.
You can learn more about the EU AI Act here:
Data Sovereignty in Business
Personal privacy is well established, as enshrined in the EU’s GDPR. But the privacy of proprietary business data is less discussed, often with great potential consequences. For example we speak with many companies that cannot use third party solutions because they are hosted in a different country, which doesn’t have the same laws as their host country (e.g. GDPR). Other times, companies don’t want to transmit their proprietary data outside of their internal networks because they are trade secrets.
At this point I take another look at the fact that all of these laws have national security exceptions. I was surprised to find that the UK’s signals intelligence service GCHQ had recently been found to be in breach of the european convention on human rights (ECHR). Maybe I missed this because of COVID, but I don’t recall it being publicised in the media at all.
My point is that governments may not have the inclination to spy on your trade secrets (although they have the capability), other companies might. Handing your data over to third parties might not be a risk you are willing to take. But of course this must be viewed on a case by case basis.
An easy solution to this problem is private AI. The fastest and easiest way to remove this risk is by building a bespoke AI stack that is customized to your needs and, crucially, inside networks you control. This provides your business with the confidence that data isn’t leaking and the flexibility to achieve your goals.
Conclusions
AI adoption is no longer optional. It’s a strategic necessity. With 70% of businesses already using AI, the real opportunity lies in leading its integration effectively.
At Winder.AI, we specialize in helping businesses unlock AI’s potential for growth, innovation, and operational excellence. Get in touch with us today to start building your AI success story in 2025.
