ISG’s 2025 study of IT budgets shows that 90% of traditional German asset-heavy industrial firms were spending 5% or less of their digital and IT budgets on AI investments. This means they are trailing far behind adjacent industries, in which investments of this kind reach anywhere up to 15%. This concerning statistic puts them at danger of not only forfeiting the global lead on AI adoption, but also of letting their competitiveness decline at an alarming rate.
The trend isn’t new – DACH is widely regarded as a region experiencing slow adoption of AI technologies. This perception of being a laggard in comparison to global markets was only strengthened at the international AI Action Summit held in Paris in early 2025.
Despite the pessimistic picture, there is a narrow window of opportunity that hasn’t closed yet. Recent public and private investments in the industrial application of AI in Germany – including the government “going all in” on AI investments and NVIDIA announcing the world’s first industrial AI cloud – are signs that the German industrial powerhouses may yet be capable of avoiding the slip into irrelevancy.
But just how have the manufacturers reacted to such news? Are they learning from peers in gaining value and, more importantly, a responsible embrace of AI – one that extends beyond ethics and the EU AI Act?
Early definitions of responsible AI were dominated by the ethics of such technology, and rightly so. This has been brought about by reports of AI applications being racist, sexist and generally not reliable.
Considering an ethical dimension brought awareness to the concerns raised from having a “black box” approach. However, as we evolve, this concept must extend further, particularly as the investments to date are not yielding the expected rewards. After all, any investment in AI is effectively taking resources from elsewhere – and not necessarily in an efficient way.
Responsible AI must grow to include questions of architecture and business viability, instead of purely matters of compliance. It should ask: am I prioritizing the right problem? Does this problem really need to be solved by AI, or could we use traditional process optimization instead, consuming less cloud resources? It should also include considerations of holistic resource investment: are there clear benefits that further the overall business strategy by contributing to the core products and services?
Please fill out this form to continue.