The issue with AI is no longer about the functionality of the technology, but about whether the individuals and processes surrounding it are ready, according to business and tech leaders at KPMG’s discussion at [Viva Tech Paris](https://vivatechnology.com).
“We recognize the potential,” remarked Nicole Buttner, CEO of Melantec and General Secretary of Germany’s MPP party. “Yet, it feels like this uneasy transition where the expectations of generative and conversational AI have not completely materialized within companies.”
Europe, previously seen as a hesitant player in the AI competition, now grapples with operational hurdles like siloed systems, fragmented data strategies, and workplace cultures resistant to transformation. As the UK aims to take the forefront, warning signals remain clear.
Despite significant investment in AI advancements, profits continue to be elusive beyond specific machine learning applications. “The genuine, scalable ROI today resides in conventional areas like computer vision and standard NLP,” Buttner articulated. “Generative AI? We simply have not reached that point in terms of enterprise influence.”
Ashish Badan, one of KPMG’s leading technology officers, compared the situation to “owning a sleek new Porsche, fully equipped…but lacking a licence.” He informed Viva Tech attendees that the concern lies in organizational infrastructure, not GPUs or models. “You require governance, systems, and people prepared to utilize the technology.”
This illustrates a wider challenge for UK firms, many of which are zealously initiating ambitious pilot projects yet encountering difficulties in moving past experimentation.
For robotics innovator Dr. Amit Andre, the merging of AI and robotics—what he referred to as “physical AI”—signals a new era of capabilities. “But with that arises a more profound question: are we applying it where it is truly needed?”
He warns against the haste to implement AI indiscriminately, underlining the risk of performative innovation, especially in publicly funded or high-pressure sectors such as UK healthcare and government.
The speakers repeatedly stressed that it is not the chips or algorithms that are obstructing AI, but rather human application. Governance, incentives, and cultural acceptance are the real chokepoints. Buttner remarked: “Numerous businesses and governments encounter the same issue of integrating diverse, incompatible systems.”
The UK faces similar decentralization issues, with numerous NHS AI pilot projects underway, but few have achieved scalability. Without cohesive digital infrastructure designed to enforceable standards, even effective solutions may become compartmentalized.
“At this moment, organizations are experiencing the coordination challenge, particularly as they deploy AI agents,” Buttner elucidated.
For additional insights, see discussions on whether the [technology itself is effective](https://www.cityam.com/nvidias-jensen-huang-praise-and-promise-are-useless-without-uk-ai-infrastructure/) and the [possible economic ramifications](https://www.cityam.com/klarna-boss-ai-will-lead-to-recession-and-mass-job-losses/).
