Recently, I had the opportunity to attend the ABBYY AI Summit 2025 in Vienna – a gathering of technologists, analysts, business leaders, and governance experts tackling one of the most pressing questions facing organisations today: how do we really adopt AI successfully?
The headline? We’re past the peak of inflated expectations – and that’s a good thing.
Confronting the Challenges of AI Projects
Dr Marlene Wolfgruber kicked off with a stark reminder: 80% of AI projects fail. Not because of the technology, but because organisations dive in without clear strategy or the foundations required for success.
Here are just some of the common pitfalls discussed:
- A lack of defined purpose or measurable outcomes
- Poor quality data and documentation
- No process transparency – businesses simply don’t understand how they operate today
- Governance gaps and privacy blind spots
- Leadership buy-in driven by FOMO, not function (60% of CEOs have invested in AI simply because of fear of missing out!)
It’s a familiar list. At Convergent, we see many of these challenges first-hand when organisations start exploring intelligent automation or AI-enhanced document management. Technology alone isn’t the answer. It’s just the tool.
Building AI That Works – For People and Process – Strategies for Successful AI Integration
Several speakers shared the same underlying message: don’t expect AI to be your saviour if you don’t understand your own business.
As Bernhard Hecker put it: “If you don’t know how your business actually works, don’t ask AI to fix it.”
That struck a chord. Our work at Convergent often begins with helping clients define and document their existing processes. It’s not the most glamorous stage of transformation, but it’s essential.
The summit highlighted several principles for building AI programmes that actually work:
- Problem clarity: Start with the business need, not the technology.
- Data quality is everything: AI is only as good as the structured, secure, and well-managed information you give it access to.
- Enhance, don’t replace: Human oversight and existing systems still matter. AI should work with what you’ve built, not bulldoze over it. Enhance what works – evolve, don’t erase.
- Invest in AI literacy: Real adoption starts at the top. Leadership teams must ‘walk the talk’ when it comes to understanding and governing AI.
- Privacy and explainability by design: If your AI system can’t be explained, it becomes a liability, especially under the evolving landscape of GDPR and the EU AI Act.
The Legal & Ethical Landscape
Privacy expert Alexander Hanff delivered one of the most thought-provoking presentations, frankly scaring the room with his breakdown of where liability sits when AI goes wrong! Hanff drove home that understanding existing frameworks like GDPR (which isn’t going anywhere) and the EU AI Act is absolutely essential before diving into AI implementation.
We also saw some truly mind-blowing insights from Resistant AI’s Fraud Detection team, showing the double-edged sword of AI in security – it’s being used to fight fraud but also by fraudsters themselves! This reality check highlighted just how complex the security landscape becomes when implementing AI in sensitive environments
Moving Beyond the Hype Cycle
As I reflect on the conversations at the ABBYY AI Summit, it’s clear we’ve moved past the peak of inflated expectations in the AI hype cycle – and that’s exactly where we need to be.
The focus has shifted from magical thinking about AI capabilities to the practical realities of implementation, governance, and sustainable value creation. Companies that understand this shift will be the ones that succeed with AI in the long run.
At Convergent, we’re committed to helping organisations navigate this more mature phase of AI adoption – where enthusiasm is tempered with expertise, and innovation is balanced with overall integration into the organisation.
The future of AI isn’t about chasing the next shiny object but about building intelligent solutions on solid foundations.
About Me: I am Nik Healy, Solutions Director and co-founder of Convergent, before which I held Management Consulting roles in Ireland and Canada. I lead the Solutions Team at Convergent and am widely regarded as one of Ireland’s leading document & information management experts. When I am not in the office, I am regularly to be found bounding down hills and mountains on my bike !
To understand how our business can help yours, contact me directly via info@convergent.ie
