AI Risk Management in the DACH region: Overviews and Trends for AI in 2024

Introduction

With AI implementations gaining traction throughout the global market, the DACH region prevails as a guiding force towards AI proliferation around the world. In this blogpost, we look to identify the trends of the region’s model adoption, coupled with the implications for the anticipated sectors of Banking and Insurance as strongly corresponding, AI-enabled industries.

‍

The Current state of AI adoption in DACH

Under the 2023 Global AI Index published by Tortoise Media (co-founded by James Harding, former Director of News and Current Affairs at the BBC and Editor of The Times), rankings on both overall and sub-sectioned metrics were identified for the DACH region. In particular, the overarching rankings consist of Scale and Intensity, while the following sub-pillars of Talent, Infrastructure, Operating Environment, Research, Development, Government Strategy and Commercial are normalized between 100 and the minimum original score.

‍

In terms of AI adoption, Germany, Switzerland, and Austria rank 8th, 9th, and 20th on an overall global scale. Uniquely, Switzerland ranks 3rd in the world for Intensity with a nominal score of 71.6, surpassing AI-powerhouses such as the United States, United Kingdom, and Canada (5th, 10th, and 7th place, respectively). Moreover, Switzerland is 4th in both the Research and Development sub-pillars. This ardor is continued by Germany’s 3rd place ranking in terms of Scale, trailing the United States and Canada. Austria hosts a strong, balanced ranking, wherein it notably places 5th on the Operating Environment sub-pillar.

‍

With this unique blend of rigor across the DACH, AI adoption has become a significant force in enterprise, especially in the banking and insurance sectors. In fact, the IDC notes that for the DACH, ICT spending will near $275 billion in 2023 as it shifts heavy focus onto AI. Moreover, it is stated that for the region, “Discrete manufacturing, professional services, and the banking industry will spend the most on AI platforms this year, while investments by the healthcare, insurance, and telecommunications industries will increase at the fastest rate into 2026.

‍

Thus, as AI models and benefits increase, such do their concerns; with wide adoption throughout the region, what are some of the driving forces guiding DACH towards AI Risk Management?

‍

Current issues with AI adoption

As noted in IBM’s Global AI Adoption Index released in January 2024, amongst the surveyed companies about 42% of enterprise-scale firms (>1 000 employees) reported implementing AI. However, these firms cited a number of adoption risks, including the top barriers of:

‍

- Limited AI skills and expertise (33%);

- Too much data complexity (25%);

- And ethical concerns (23%).

‍

Moreover, on a regulatory note, considering the concentration of banking and insurance firms in the DACH region, Item 3 is especially proliferated as high-risk EU AI Act applications prevail in the area. Such is only exasperated by the need for trustworthy AI across all firms - with IBM identifying that 44% of respondents are looking to develop ethical AI policies.

‍

Rate of adoption of AI Risk Management mitigation

IBM continues to note that though numerous companies with deployed AI are facing significant barriers in the process, “well under half” report they are taking the necessary steps towards Trustworthy AI, such as:

‍

- Reducing bias (27% of respondents);

- Tracking data provenance (37%);

- Or making sure they can explain the decisions of their AI models (41%).

‍

The DACH has been instrumental in pursuing early AI risk management concepts from top-down approaches such as the EU AI Act, while firms also begin to introspectively search for optimization and risk management solutions. Switzerland’s Artificial intelligence and international rules Report, highlights the key issues faced by enterprises and their relative legal applications:

‍

- Explainability: the “black box” nature of AI that calls for transparency and traceability in model construction;

- Bias: adjusted fairness for algorithms to be free of discriminatory behavior, whether it be data or model-related;

- Surveillance and Manipulation: the question of individual rights for AI systems’ interference with privacy under human rights rules;

- Accountability and Liability: in this context, involving the liable figure when AI models suffer incidents leading to harm;

- Life Cycle Regulation: continuous oversight of models—especially ML, as noted by the FDFA—which ensures safe use of AI throughout its lifecycle.

‍

These points reflect the intense actualities for DACH firms as they search for AI risk management solutions in spite of the mounting issues for AI-powered enterprises.

‍

At Calvin, our expertise lies in guiding firms towards technical, ethical, and regulatory AI excellence. To aid in compliance, guarantee unbiased algorithms, enhance the efficiency of company AI portfolios, and much more, we adopt a quantitative approach to AI risk management to ensure your models remain compliant and optimized throughout their life cycles.

‍

Let’s make trustworthy AI a tangible reality for the DACH region and beyond — schedule a demo with us today!