The Reinvention of Invention

AI-Assisted Invention: From Pioneering Research to Industrial Reality

From the first prototypes to thousands of patents. What happens in the next five years?

Thursday 5 November 2026 | 16:00 GMT / 17:00 CET / 11:00 EST

Invention is where a company decides what its future products, capabilities and competitive advantage might look like. Yet for much of industrial history, the process of creating inventions has remained comparatively unpredictable and difficult to scale.

When Iprova began developing AI-assisted invention technology in 2012, the idea that machines might help human inventors identify unexpected technical connections was far from mainstream. There was no ChatGPT and no current generation of large language models. The early task was to see whether AI could reveal possibilities that even experienced inventors might overlook.

The work began to attract attention. New Scientist explored how creative computers could develop ideas beyond the reach of human imagination. In 2017, Switzerland’s Neue Zürcher Zeitung profiled Iprova’s founder under the headline “Der britische Turbo-Entwickler aus Lausanne”. In 2020, BIC and Iprova launched the world’s first data-driven invention lab, taking the approach into an industrial setting.

The technology has since contributed to inventions behind thousands of patents and patent applications. Granted patents resulting from this work have been cited by all ten of the world’s largest technology companies. Iprova has demonstrated efficiency improvements exceeding 20× in defined invention workflows. Its approach was examined by IEEE Spectrum in 2024, while Iprova’s leadership co-authored a 2026 Harvard Business Review article on what it takes to translate AI into business results.

That track record matters. But the most consequential changes may still lie ahead.

On Thursday 5 November, we bring together three perspectives on the development of AI-assisted invention, its origins, its present capabilities and what advances in AI could make possible over the next two, three and five years.

PAST. Could AI help humans invent?

Professor Patrick Ruch | HEG-Genève, HES-SO

Prof. Patrick Ruch helped develop Iprova’s original research prototypes. He will revisit what those systems were designed to achieve, what was technically possible at the time, and what the early work revealed about using machine intelligence to connect knowledge across disciplines.

Which assumptions proved right? What took longer than expected? And what can today’s AI researchers learn from the first generation of invention systems?

PRESENT. What can AI-assisted invention do today?

Dr. Chris Wright, CTO of Iprova | Live technology demonstration

Fourteen years on, AI-assisted invention is no longer an academic proposition. Enterprise teams can use the technology to investigate opportunities, make connections across technical domains, create inventions and evaluate their potential.

Our Dr. Chris Wright, CTO of Iprova will demonstrate the current state of the art, following a real invention workflow rather than simply showing a succession of software features. The focus will be on what an inventor can actually achieve today: exploring technical possibilities, developing inventive concepts and bringing human expertise to bear on their commercial and patent potential.

We will also examine what has changed in practical terms: the speed of invention, the breadth of knowledge that can be brought into play, and the increasingly important role of human judgement.

FUTURE. What will be possible in two, three and five years?

Dr. Catherine Breslin | AI researcher, former Amazon Alexa scientist, founder of Lichen AI

AI is progressing rapidly, but forecasting its impact on invention requires more than extrapolating from the latest model release.

Dr. Catherine Breslin will consider which advances in reasoning, agents, multimodal AI and scientific models may matter most. Could systems begin to identify invention opportunities continuously as new knowledge emerges? How much more of the inventive process might they support? And what will remain fundamentally dependent on human expertise and judgement?

She will examine realistic possibilities over a two-, three- and five-year horizon, including the limitations, uncertainties and practical barriers that still have to be overcome.

The discussion will be moderated by Julian Nolan, Founder & CEO of Iprova, who will bring together the speakers’ perspectives and lead the conversation on how AI-assisted invention has evolved, where it stands today and what the next five years could hold.

THE BIGGER QUESTION. What should R&D and IP leaders do?

We will conclude with a panel discussion and audience Q&A focused on the competitive consequences. If the cost and time required to create inventions continue to fall, what will distinguish companies that create real value from those that simply produce more ideas?

Faster invention is valuable. But the ability to direct invention towards the right problems, anticipate future needs and build defensible positions around the results may be considerably more valuable.

Join us for a look back at the origins of AI-assisted invention, a live demonstration of what is possible today, and an informed discussion of what comes next.

Thursday 5 November 2026 | 16:00 GMT / 17:00 CET / 11:00 EST