Security of AI systems as a business booster
Security of AI systems as a business booster
25 October 2023
In the future, NTT DATA will use the AI Quality Framework developed by TÜV SÜD to inspire greater trust in applications that use artificial intelligence (AI) and to assure their quality. The cooperation agreement covers use of the AI Quality Framework developed by TÜV SÜD as well as appropriate analysis tools, enabling the customers of NTT DATA to design and develop high-quality market-focused AI applications.
The TÜV SÜD AI Quality Framework provides organisations with guidance and tools that enable them to identify any potential risks involved in their AI applications and test their applications for fairness, explicability and robustness. In addition to the above, IT service provider NTT DATA also supports its customers in implementing actions to improve and assure AI quality and a machine-learning operations (MLOps) lifecycle – which ensures that there will be no performance degradation over time. “The pressure on companies to transform their operations is increasing, with a growing number opting to implement AI systems to automate their processes. The TÜV SÜD AI Quality Framework offers our customers guidance to prepare for the upcoming legislative requirements of the EU. Through this support, the requirements will not become a constraint but will instead foster trust and security in automated operational processes”, explains Stefan Hansen, CEO and Chairman of the Board of NTT DATA DACH.
The AI Quality Framework supports organisations seeking to develop or use quality- and market-focused AI applications. The framework is based on currently applicable global regulations and standards and takes in the cornerstones of high-quality AI applications, including data management, AI robustness, life-cycle and quality assurance of AI applications, safety and security, and control and testing of AI applications. Customers can use analysis tools to test the fairness, explicability or robustness of their AI applications. “By systematically applying the AI Quality Framework as a basis, we enable organisations to design and develop high-quality AI applications in a time-efficient and targeted manner, either independently or with the assistance of third parties. At the same time, the framework helps companies to keep track of regulatory trends and approaches, and thus to develop marketable applications”, emphasises Dr Andreas Hauser, CEO Digital Service, TÜV SÜD.
A study by the German Economic Institute examining how Germany will benefit from intelligent technologies estimates that generative AI could contribute as much as EUR 330 billion to Germany’s gross value added in future. It further states that roughly 17 per cent of German enterprises are already using AI in areas such as text, data analysis and marketing.
At the same time, consumers are calling for a legal framework to ensure safe and secure AI applications. A representative survey of the German population carried out by TÜV Association (TÜV-Verband) asked for views on legal requirements addressing the issue of AI. The results were very clear: 91 per cent of respondents called for legislation to establish a legal framework for ensuring the safe and secure use of AI. In the opinion of 83 per cent of respondents, European values should be taken as orientation for the regulation and use of AI.
However, AI-related risks are frequently underestimated. Algorithmic bias often results in decisions that disadvantage certain groups by discriminating against them on the basis of age, gender or ethnicity. A problem that not only causes loss of trust among customers, but also violates EU values and laws and may be sanctioned by high fines in the future. To minimise the negative impacts of AI applications, the EU’s AI Act, which will set out rules and guidance for the use and development of AI systems, is likely to see finalisation later this year.
NTT DATA is a globally operating provider of business and IT solutions and has years of experience with the topic of AI. The IT service provider supports organisations with compliance with the stricter EU regulations on AI use and development. In addition, customers can access a white paper offering insights on the topic of “Trustworthy AI” on NTT Data’s website.
Global testing, inspection and certification (TIC) company TÜV SÜD offers a host of services related to AI application quality management (QM) and quality assurance (QA). In addition to training, TÜV SÜD provides quality analysis, risk assessments and audits. Further information is available at: https://www.tuvsud.com/en/themes/artificial-intelligence
Press-contact: Heidi Atzler