In May, we hosted the AI-Assisted Invention Summit in Lausanne. The room included senior figures from Microsoft, IBM, Sony AI, Dassault Systèmes, Capgemini, STMicroelectronics and others working at the frontier of AI-assisted invention. It was a frank, high-quality conversation and five themes kept emerging.
What follows is a summary of each, along with what we think they mean for R&D and IP teams today.
1. Access to AI is no longer the competitive advantage. Designing work around it is.
Every organisation now has access to AI. The question is no longer whether to adopt it – that debate is over. What distinguishes the teams pulling ahead is how they have redesigned their invention processes, decision-making and team structures around AI. Who does what. Where AI leads and where the human inventor does. How invention fits into the wider AI-enabled stack.
This is an organisational design challenge as much as a technology one, and most organisations have not yet taken it seriously.
2. The most effective AI-assisted inventors are not simply faster. They are more deliberate.
Speed is the most visible benefit of AI in invention workflows, and it is real. But the teams producing the strongest results are the ones who have been thoughtful about the division of labour. AI processes signals at a scale no individual can match. It does not decide which direction is worth pursuing. That judgement belongs to the inventor and the best inventors treat it that way.
3. Inventor judgement, creativity and domain expertise are becoming the real competitive edge.
Here is the finding that surprised most people in the room: as AI takes on more of the analytical and procedural work in invention, human expertise becomes more valuable, not less. The creative leap, the cross-domain instinct, the ability to identify which technical direction will produce defensible, commercially significant IP – these are increasingly scarce and increasingly consequential.
The organisations that understand this are investing in the conditions that allow their best inventors to operate at that level. The invention layer is re-emerging as a genuine competitive advantage.
4. AI’s greatest impact comes from how organisations are structured around it – not from individual performance.
Every organisation has a few people who are exceptional at working with AI. That is not a strategy. The organisations achieving durable gains are the ones that have built AI capability into how their teams work – shared platforms, clear governance and integrated workflows that belong to the organisation rather than a handful of individuals.
5. The inventor’s role does not disappear with AI. It moves up the stack.
This was the insight that landed hardest in Lausanne. When AI handles signal detection, prior art screening, evaluation and documentation the work that absorbs a significant portion of an inventor’s time – inventors are freed to operate at a higher level. Strategic direction. Cross-industry creative thinking. The deep domain knowledge that turns a technical observation into genuinely valuable IP.
AI compresses the routine. It creates the conditions for what only an experienced inventor can do.
What this means for Invention Studio 3 (IS3), Iprova’s AI-powered invention platform
None of these insights surprised us, because they are the principles IS3 was built around. The platform gives teams near real-time signal detection, automated documentation and a complete end-to-end workflow, so inventors can operate at the level the Summit described. It is built for organisations, not individuals. And it creates a clear, auditable record of where AI contributed and where the human inventor decided.
We are sharing all five insights in depth at our webinar on 1 July with a live IS3 demonstration and a panel that includes Prof. Frank T. Piller from RWTH Aachen University and Joff Wild, former editor of IAM, as moderator, Dr. Harry Cronin and, Jon Foster from Iprova.
Date: 1 July 2026 | Time: 15:00 BST | 16:00 CEST | 10:00 EDT | 60 minutes
Register here.