Digitalisation and data ethics go hand in hand at PFA

At PFA, we are improving our digitalisation to meet future requirements for a modern pension company 
and to make PFA even more relevant for our customers. 
PFA will increasingly be using new digital technologies to strengthen our general service and advisory services, increase self-service options, streamline case processing and boost the customers' health – for the benefit of our customers, their workplaces and society in general.

AI has great potential for improving the customer experience

Today, PFA’s customers have access to a modern online platform, My PFA, which they can use to get an overview of their pension plan, receive recommendations, make simulations of future pension payouts, make changes via self-service and to monitor PFA’s case processing. 

Thanks to advanced data processing and AI (artificial intelligence), in the coming years PFA will be able to offer you new digital services that will be far more targeted towards individual customer needs and circumstances. And with this new technology, we will also be better able to identify customer groups that may particularly benefit from PFA’s offers concerning preventive initiatives and treatment. Thereby expanding the safety net provided by PFA’s insurance plans. 

We take our ethical responsibility seriously

PFA knows that increased digitalisation comes with a great responsibility. We want to ensure that our customers feel safe at PFA, regardless of how comfortable they are with digital solutions, and we want to ensure that AI is only used for purposes that are compatible with PFA’s fundamental values as a customer-owned pension company. 

The responsible use of data is not just about complying with the general regulations concerning the protection of personal data that PFA is subject to, it also involves PFA’s own data ethics principles. These serve as the compass for PFA’s ongoing development and our advanced use of data and AI.  

Today, there are no fixed definitions when it comes to data ethics, and the perception of good data ethics is constantly changing. The same applies to PFA’s work with data ethics and data ethics principles. PFA will continue taking an active role in the important societal debate on good data ethics and implementing concrete initiatives aimed at ensuring a deeper implementation of data ethics in PFA’s operations and among its employees. 

Data ethics principles

PFA has decided upon the following data ethics principles for the use of AI in customer solutions:



1. AI must be used on a fair and responsible basis and be in the best interests of our customers, the community of our customers and society.

AI is a technology that is under significant political attention, both in Denmark and internationally. There are several initiatives in progress, which are aimed at ensuring that this technology will be used on a responsible and sustainable basis. PFA is closely monitoring those developments, and we will continually be implementing recommendations and statutory requirements in our work with AI. 

Learn moreView less

At PFA, we are convinced that a fair and responsible use of AI can easily co-exist with the ambition of digitalising our operations and customer services. For us, the critical factor is that AI should be used to serve the best interests of our customers and and that the considerations for the individual customer, the community of customers and/or society are made with due regard to a specific balancing of interests.  

PFA is increasing the use of AI in the following main areas:

Operations and case processing: Internal tools that contribute to the optimisation of internal processes, task management and responding to customer inquiries

Customer advisory services: More customer-friendly information and advisory services at My PFA, targeted towards the individual customer’s situation and needs

Prevention: More precise identification of target groups that would particularly benefit from PFA’s offers concerning preventive initiatives and treatment to avoid long-term sick leave and loss of occupational capacity

PFA does not use AI to determine individual prices and terms or to make automated decisions on whether customers are entitled to payouts. PFA also does not collect health information from the customer’s mobile devices on an ongoing basis (smartwatches, health apps, etc.).  



2. The development of algorithms for AI must live up to recognised technical requirements and recommendations and follow a controlled and documentable process with an integrated assessment of whether the results are explainable and non-discriminatory.

PFA has formal procedures to manage, control and document the development of algorithmic models. This ensures that all data ethics requirements for algorithmic models are complied with, that the models are explainable and do not discriminate or otherwise make inappropriate decisions.  



3. Transparency and information must ensure understanding of and trust in PFA’s AI-based digital customer solutions.

As PFA in the coming years begins to increasingly use AI for more customer-facing functions and services, the need for customer information will grow proportionally. 

It is critical for PFA that customers are always informed if they receive advice, guidance or decisions that are based directly on AI. The customers must know that they can contact PFA’s employees if they are in doubt or need specific advisory services and guidance. 





4. PFA’s customers must always have access to advice and guidance and have their complaints processed by a physical person.

PFA is a customer-owned pension company, and it is deeply rooted in our values that we want to be here for our customers and be close at hand. That will remain the same in the future. 

Increased digitalisation and automation will undoubtedly change the PFA customer experience, and overall, it will also change the need for contacting PFA. However, there will always be situations where our customers will need to talk to a PFA employee, for example, when they are ill. PFA’s employees will continue to be available with advisory services and guidance precisely as they are today.  

Account of data ethics

In our account of data ethics, we describe among other things which types of data PFA processes, our partnerships with external parties, our considerations related to new technologies, data-ethical dilemmas, awareness training and our ongoing evaluation. The account of data ethics forms part of the management’s report for PFA Holding’s and PFA Pension’s Annual reports.

Data security

Data ethics is also about data security. 

As a pension company, PFA is used to and has set up its operations to process a lot of customer data. We are subject to comprehensive statutory obligations for data security and data protection, and processing data securely and responsibly is an integrated part of our culture and processes. At PFA, IT systems and processes are developed with a focus on our customer’s data security, and we have strict requirements for the validation and quality of the data that we use in the customer relationship.

PFA has organised itself in such a manner that there is a continual monitoring and control to ensure that data security remains robust. We have procedures that ensure that any suspected or identified breach of data security is immediately investigated so that the scope and impact of the incident on PFA’s customers can be determined, and the Danish Data Protection Agency is also notified as per the applicable regulations. 

PFA collaborates with relevant authorities and industry organisations to increase the level of data security and resist the threat of cyber-attacks.