Many organisations are engaged in digital transformation, with algorithms and other advanced analytical solutions increasingly being part of operational management. For executives and managers, it is very important to be able to rely on the insights, decisions and actions performed by these analytical solutions in order to take full advantage of technologies such as Artificial Intelligence (AI).
AI in Control is part of KPMG Trusted Analytics: services aimed at implementing data analytics and AI in a controlled and managed way, without making any concessions on innovation. With Trusted Analytics, KPMG carries out independent reviews, exposes actual or potential vulnerabilities and risks and points the way to solutions. In this way, we offer organisations assurance, create trust in Artificial Intelligence and make it possible to utilise the full potential of AI.
AI in Control is one of KPMG's Non-financial Assurance services, through which we help organisations to keep a grip on progress.
AI in Control, what exactly is it?
AI in Control has four pillars with which we can support your organisation. We tailor the services to you, so that they seamlessly link in with your organisation and your AI solutions:
Good risk management calls for insight: where in your organisation are analytical solutions used, what is their impact and what risks do you run? We can help you answer these questions. With our easy to apply quick scan, the Analytics Risk Index, we quickly identify where your analytics and AI solutions are used and which potential risks you run as a result, with the help of a workshop and simple questionnaires. This knowledge will enable you to take targeted steps to improve your control environment.
Changes to existing guidelines and the introduction of new legislation and regulations also call for insight: insight into the nature of the changes and what they mean for your organisation. We provide that insight. For instance, for one organisation we investigated how the new guidelines would affect existing AI solutions, what the risks would be if these guidelines were not complied with and what the costs of bridging the gap would be. This also gives you insight into your analytical solutions.
The use of analytics and AI comes with new risks. In some cases you will be aware of them, in others not. We can help. Based on a detailed risk assessment, we will identify which risks you are exposed to. We will then use our experience, scientific knowledge and library of tools to help you control those risks. Starting from your existing risk management approach, we will help you to implement new measures. We are experienced in that role. For instance, we have developed various testing frameworks for algorithms. One of those frameworks was focused on enabling an organisation to evaluate whether the algorithms it was using were fair and transparent.
Utilising the opportunities of AI while controlling the risks calls for governance and policy. What does that mean in practice? Based on a governance maturity scan, we will help you answer that question. We will help you overcome the identified shortcomings, both within processes and in organisational terms. We have experience of supporting organisations in implementing ethical frameworks for data-driven decision-making. With our starter kit, we enable organisations to implement the framework. We also offer training to help employees put the framework into practice effectively. Helping deliver governance within your organisation that utilises the opportunities of AI in a reliable manner.
Do you know for sure that a specific algorithm is reliable, explainable and ethically responsible? We provide independent assurance with our technical deep dive and review. For example, an algorithm for predicting remaining healthcare costs. We evaluated it by examining the development process, the algorithm and the output from a technical and financial perspective. Are you more interested in knowing whether your algorithm complies with specific requirements? For example, whether it is fair? Or secure? We do that too. Whichever form you choose, ultimately you will receive answers to two essential questions from an independent party: does the algorithm do what it is supposed to do and will it continue to do so in the future?
KPMG has a long record in the area of IT auditing, Non-financial Assurance and risk management. The core team behind AI in Control consists of people with diverse expertise: from data scientists and IT auditors to privacy experts, legal specialists and ethicists. Moreover, with over 155,000 colleagues in 144 countries, we always have the right expertise on hand to match your specific requirement.
Is your company changing due to the use of AI? Are you experimenting with new AI solutions or do you want to implement them into your existing business processes? Regardless of what your AI challenge is, with our four pillars we can help make sure your AI is in Control.
Would you like to know what KPMG can do for you in the field of AI? Please contact Frank van Praat, director AI in Control.