Finansforbundet insists on responsible development of AI

We acknowledge that AI may free up time and increase efficiency. But efficiency should never become a goal in itself. We are working to ensure that AI is used to reduce the workload and free up time for core tasks – not to lay off employees or increase the pace.

AI should be a driver of increased quality over quantity. AI should contribute to sustainable working lives and a good working environment, reduce stress and burnout.

Therefore, we work for:

• AI constitutes a labour market issue the benefits of which we should reap together
• AI should not lead to increased stress and control – systems must be transparent and not take away employees' autonomy
• Protection against surveillance and unfair use of employee data must be a fundamental requirement – not an afterthought. 
• We need to find solutions for how to increase employees' rights and control over their own data.
• AI should support skills, not replace them.
• AI systems must be transparent and fair to facilitate the understanding and challenge of decisions.

AI must not shift the balance of power in the labour market, but should instead support the Danish Labour Market Model, according to which the labour market parties collaborate on finding solutions.

Finansforbundet believes that organisations should develop clear and practice-oriented strategies and guidelines for the use of AI with a view to setting and communicating goals and a direction for the organisation's technology use. That will create security, establish a common direction for all stakeholders in a company and ensure a shared understanding of what is good and what is undesirable AI practice. These clear guidelines should also include procedures for how to continuously capture and manage unintended incidents expected to emerge with the use of AI. 

Therefore, we work for: 

• All companies should develop AI strategies clearly communicating the company's visions and directions for AI use to ensure that everyone understands them and may adapt accordingly. 
• Clear guidelines and communication, so everyone knows the rules of the game. 
• Especially where AI is used for HR and management – for example during recruitment, performance assessment or decisions on salary and development – clear guidelines and communication should be established in terms of what is collected and how, what is used and how (the AI model), for what purpose and who has access.

Unfortunately, we see that decisions about AI are made without involving employees or delegates. We need to change that. AI should not become a black box controlling our jobs without being rooted in democratic principles. Allowing employees and their representatives to be involved not only leads to better solutions; it also builds trust and increases the quality of implementation. It is, simply put, in everyone's interest.

Therefore, we work for:

• Inclusion of the employee perspective – involving employee representatives as the collective representative of the employees will ensure actual influence. Ensuring that union representatives are consulted in a timely manner, granted access to relevant knowledge and have the opportunity to influence decisions.
• Actively involving managers and employees with professional expertise and experience. They are the ones who know the tasks, workflows and customers – and, therefore, they are the most qualified at assessing how AI should be used in practice.
• Putting in place fixed procedures and frameworks to ensure employees' rights and opportunities for influence. The right to be heard and involved should not depend on the goodwill of management – it should be part of the approach we adopt to technology in the sector.

We do not want detailed regulation, but the market does not automatically secure accountability. Therefore, we are fighting for political regulation that protects employees but also contributes to creating fair competition. When companies do not have common rules, we risk that accountability becomes a disadvantage, and that those who lead with ethics and data responsibility are overtaken by those who cut corners. 

We believe that common and binding rules are crucial for creating trust and sustainable competition in the sector. It could and should be a competitive advantage to take data ethics, employee involvement and long-term values seriously.

Therefore, we work for:

• An EU directive on algorithms in the workplace that ensures participation and human control.
• A reintroduction of the AI Liability directive to keep employers accountable for mistakes made by AI.

To ensure a balanced labour market in which no one is left behind, everyone should have equal opportunities to learn about and apply the technology. Today, it differs greatly who actually uses AI and who feels equipped to use it. Significant differences exist with respect to gender, age, educational level and company size. Far too many employees do not have access to the tools or the knowledge to get started – or feel safe using them.

Therefore, we work for:

• Combining access to additional training in AI with an appreciation that learning is not only acquired through books. Employees are to a higher degree developing their skills when they get the opportunity to experiment with and test AI tools as part of their day-to-day tasks. Therefore, both time and space must be allocated during the work day for employees to learn by doing and enhance their skills as they carry out tasks.
• Equipping managers and delegates to navigate (use effectively, be critical, etc.) a workday in which AI is used as a work tool or management tool for assessment and recruitment of employees, for example.
• Providing everyone access to AI tools and the time to try them out and learn how to use them in practice.

AI holds enormous potential to strengthen Denmark's competitiveness, create jobs and increase productivity. To ensure that we make the most of the opportunities – in a way which benefits both businesses, employees and society – AI should be developed and used responsibly based on clear ethical guidelines.

Therefore, we work for:

• Using AI as a lever to gain more time for core tasks, allowing employees to focus on value-adding work instead of routine tasks.
• Increasing collaboration and investment in Danish and Nordic fintech companies.

In an uncertain world, we need to strengthen the European digital sector, including AI sovereignty, and ensure that AI is not monopolised.

Therefore, we work for:

• A collaboration between Denmark, the Nordic countries and the EU to make responsible AI and data ethics a competitive advantage – rather than allowing regulation to be perceived as a barrier to development.
• Greater investments in public and ethical AI infrastructure.
• Securing storage of data in European cloud services.

Trade unions also have a responsibility to contribute constructively to the development. We should not simply take on a reactive role responding to the initiatives of others. We also need to come up with solutions. Finansforbundet takes its responsibility seriously.

We are actively working to ensure that the labour movement:

• Finds a more powerful shared voice in the AI debate (for example through the Danish Trade Union Confederation)
• Participates actively to influence Danish and international forums on digitalisation
• Makes sure that delegates use their voices and influence – and possess the right qualifications to do so
• Takes part in establishing a national AI committee under the Danish Ministry of Employment, which will monitor the impact of AI on working life – putting  employees at the centre.