Responsible AI

Managing algorithm risk requires perspective: where in your organization are algorithms being used, what is their impact, and what risks are you facing? We can help you answer these questions. With our accessible quick scan, the Analytics Risk Index, we use a workshop and simple questionnaires to quickly map out where you are using algorithms and the potential risks you run in doing so. This perspective enables you to manage the risks you may face with algorithms in a targeted manner.

Changes to existing guidelines and introductions of new laws and regulations also require perspective: perspective on the nature of the changes and what they mean for your organization. We bring you this perspective. For example, we researched for an organization how new guidelines affect existing AI solutions, what the risks are if these guidelines are not met, and what the costs are for overcoming these deviations.

Increasingly, organizations face the challenge of demonstrating that they are using data-driven applications responsibly. Customers, regulators, civil society and management itself need assurance about the crucial processes in which algorithms are used. This goes beyond the question of whether models generate sufficiently reliable outcomes. For example, organizations must also be able to show their stakeholders that their data-driven applications are compliant with laws and regulations (such as the AI Act and DSA/DMA) and free from undue discrimination and bias. This raises several questions:

• What criteria are relevant to the audiences involved?
• What (audit and assurance) standards apply?
• What level of assurance can we provide?
• When are relevant processes sufficiently demonstrable?

KPMG Responsible AI has extensive experience in performing audit and assurance engagements in the domain of (complex) data analytics and AI. Based on our practical knowledge, we offer sound advice on the possibilities and limitations of auditing data-driven applications. Efficacy and thinking from the perspective of the actual risks are paramount here. With our unique combination of knowledge in the field of auditing and data science, we are able to perform audits that go beyond a process-oriented test, and provide insights 'under the hood' of your most critical data-driven applications.

Algorithm governance and risk management are the set of frameworks, policies, processes and measures that together ensure the responsible deployment of algorithm applications. We support your organization in developing and setting this up using the nine dimensions in our governance framework. Based on the maturity of your analytics, algorithms and AI on the one hand and governance structure on the other, we support you with the implementation of these dimensions, or a subset thereof. We always consider these elements in conjunction with and ensure their embedding in the existing way of steering, controlling and overseeing the organization.

Data ethics is an increasingly important topic for organizations and in society. At KPMG, we believe that the ethical use of data and artificial intelligence is not a luxury, but a prerequisite for the responsible and effective use of technology in today's digital age.

Why Digital Ehtics?

Algorithms can produce inequalities, violate privacy and limit individuals' autonomy. Organizations today are expected to make explicit choices regarding ethics. For example, what trade-off do they make between privacy, non-discrimination and security? More and more organizations are aware of this, but trust does not come naturally. In the context of digital ethics, these are some of the services we offer: 

• Training and awareness — Our experts provide training on Digital Ethics and help you to appropriately create awareness and knowledge about Digital Ethics in the culture of your organization.
• Ethics by design — Prevention is better than cure. We help you to develop data and algorithms ethically, without hindering innovation.
• Ethical Risk Assessment — Which algorithms are actually used within your organization and which risks are you running? We map your biggest risks and opportunities and offer concrete tools to mitigate and prevent risk.
• Digital Ethics Framework — We help you create a vision of Digital Ethics and a strategy to roll it out. You can think of drafting ethical principles, policies, roles and responsibilities and setting up an ethics committee.
• Digital Ethics Audit — You have already taken steps in the area of Digital Ethics, for example, drafting ethical principles. But are you complete, and are ethics being carried within the organization? We take an independent look at how expectations and practice match and offer improvements.