AI in Luxembourg's industry: an evolution, not a revolution

At the Scynergy conference, held at the Luxembourg Chamber of Commerce on 14–15 April, industry leaders from ArcelorMittal, MixVoip, IEE and Vodafone debated AI's real-world impact on their sectors. Their consensus: AI adoption in Luxembourg's industry is a pragmatic evolution, not a revolution — and significant hurdles remain, especially for SMEs.

AI as a productivity tool: hype vs. reality

The initial hype surrounding generative AI is giving way to a more measured reality in Luxembourg's industries. For many, AI is a powerful tool for efficiency rather than a disruptive revolution. "For me, the truth is three years after ChatGPT, we use it to write better and faster emails... and that's basically it," stated Loïc Didelot, CEO of ICT services company MixVoip. He noted that while developers are coding faster, the widespread industrial applications are still missing.

Alain Schumacher, CTO of automotive supplier IEE, agreed, emphasising that AI's integration has been a gradual evolution. "We did not wait until Mr. Altman came," he remarked, referencing the long-standing use of technologies like fuzzy logic in their sector. For him, the focus must extend beyond the hype to practicalities. "We have to look at... the business case, because what we have not talked today is... energy... licenses... putting system in place which are safe," Schumacher stressed, highlighting the significant and often underestimated costs of infrastructure and maintenance.

From AI-assisted productivity to industrial automation

In larger corporations, AI is already a cornerstone of strategy. Mihaela Ambrozie, Head of Operations at Vodafone, explained their journey from assisting employees to leveraging AI for data-driven decision-making. While the goal is to increase productivity, Ambrozie clarified that this doesn't necessarily mean job cuts. Instead, "there might be a shift... to hire different type of skills," she noted, framing AI adoption as a catalyst for workforce evolution. This journey is leading toward automation, though she stressed, "we want the human to be in the middle."

Similarly, at ArcelorMittal, AI is delivering tangible results in heavy industry. "We use it to optimize energy consumptions in furnaces where 1% of gain means massive savings and lower emissions," explained Maissa Sghaier, a research engineer at the steel giant. She highlighted the importance of grounding AI in physics to ensure models are reliable and trusted by operators. For Sghaier, this is the path forward, where AI "will allow us to move from local incremental optimization to continuous data-driven improvement at scale."

Why SMEs struggle with AI adoption

A recurring theme was the significant barrier for small and medium-sized enterprises (SMEs). "Not enough structured good data," stated Mr Didelot, is the primary obstacle preventing SMEs from leveraging AI. He painted a challenging picture for the near future: "The big enterprises... will leverage AI... but for the small and medium companies, it will become very tough." This challenge of data readiness and the lack of in-house expertise is a critical hurdle for a large part of Luxembourg's economy.

The panel's consensus was clear: while we may overestimate the change in the next two years, the transformation over the next ten will be profound. Navigating this shift requires overcoming the exact hurdles they identified, from the data readiness gap highlighted by Loïc Didelot to the need for new skills mentioned by Mihaela Ambrozie. For Luxembourg's businesses, particularly SMEs, succeeding in this new era will depend on getting these foundational elements right—a challenge that national support initiatives like the Luxembourg AI Factory are designed to address.