The insights shaping AI in 2026
Three days on the ground at VivaTech 2026
This year, something shifted. The hype and curiosity around the first generative AI demos have gone. On the Microsoft stand, running back-to-back talks and live demos with the Hitachi Solutions team, one thing became clear: no one is asking what AI can do anymore. Everyone wants to know how to make it work in practice.
The paradox? Investment is pouring in, yet the vast majority of projects are still stuck at the pilot stage. According to McKinsey, $40 billion has been invested in enterprise GenAI, and yet 95% of pilots deliver no return. Only 5% of companies realise AI value at scale and they are growing five times faster than their peers.
Why? It’s not a technology problem. It’s a challenge of adoption and scaling.
Three key lessons from our experts on the ground:
- Adoption must be built in from day one
During our session with Club Med, one insight stood out: scepticism is fading, but the critical question remains — where do you start, and which path do you take? Whether it means deploying a high-impact initial use case to bring the organisation on board, or building the governance framework first, the journey looks different for every company.
“AI’s capabilities are no longer in question. Resistance has largely disappeared — everyone is moving forward. The real question is where to start and which path to take.” – Clément Grammont, Pre-Sales Director at Hitachi Solutions
As our work with Hitachi Rail showed, anticipating human and organisational adoption from the outset is the one non-negotiable condition for turning investment into real value. The results speak for themselves: 86% of participants now feel confident using Copilot effectively, with 48 Champions engaged across 15 countries and 174 HR colleagues trained.
- Industrialising agentic AI while managing risk
Autonomous agents captured visitors’ imagination this year. But moving from demo to deployment raises complex questions: how do you test and guarantee the quality of systems that are inherently unpredictable?
During our session on our work with the Environment Agency, we shared our conviction: even if you cannot test every prompt, a robust platform, reusable capabilities and a fit-for-purpose quality assurance framework enable you to manage risk effectively. Governance is not a brake on innovation — it is what makes innovation sustainable.
- AI maturity is a collective challenge, not an individual one
On the stand, our teams observed that AI maturity cannot be measured at an individual level — which is more about knowledge and skills. What matters is organisational maturity.
We often see a gap between junior profiles pushing the boundaries of what models can do and senior leaders who are, rightly, more cautious — focused on data security, reliability and alignment with business processes. The role of a partner is to bring these perspectives into alignment.
Turning experimentation into lasting value
Generating interest at a trade show is just the first step. The real value Hitachi Solutions brings begins where the demo ends: helping organisations structure their AI journey, build secure foundations and turn AI into measurable business impact.
We brought this to life through several end-to-end demonstrations:
- Customer Experience: Deploying real-time voice agents and Copilot on Dynamics 365 Contact Centre to automate the resolution of complex customer requests.
- Operations & Sales: Automating the order management cycle through email classification and data extraction with Copilot Studio.
- Project Management: Using AI to identify scheduling risks on critical projects and suggest remediation scenarios in real time.