These Principles are guided by our purpose as a company, our core values, and our commitment to sustainable and responsible business conduct. These Principles apply to the purchase, development, implementation and usage of AI systems across the Adecco Group. As such, these Principles add to but do not replace existing policies on issues such as Information Security, Data Protection, Intellectual Property, and the Group Code of Conduct.
We are active participants in discussions on AI in society, particularly in the context of the world of work, and we work closely with policymakers to support regulation that enables responsible use of AI.
Our end goal is to achieve a Responsible use and development of AI within the Adecco Group.
Responsible means that we adhere to 5 Principles:
Our use and development of AI technologies will be:
Ethics is at the heart of who we are and is the overarching principle in our approach to the use and development of AI technologies.
We recognize the grave consequences of bias, ranging from the marginalization of vulnerable communities to reinforcing prejudice and discrimination.
To counter this, we apply as well as advocate for inclusive AI practices that prioritize diversity and accessibility, ensuring the involvement and consideration
of relevant stakeholders across the entire lifespan of these technologies.
We aim for both data equity (input) as well as output equity with data equity including both representation and feature equity and output equity, striving for impartial and fair results. Whenever data equity cannot be ensured (e.g. due to technological or other constraints), outcome equity will be controlled even more rigorously. We always aim to have a human in the decision-making loop, particularly in AI systems used in HR processes such as recruitment or selection, promotion, termination of employment, for task allocation and for monitoring and evaluating performance.
We ensure that the AI systems we develop and use will neither cause nor exacerbate harm. This entails the protection of human dignity as well as mental and physical integrity. AI systems should benefit all human beings, including future generations. We actively seek to take into account intercultural and language differences in using, developing and assessing AI systems.
We will thus ensure that AI systems we deploy and utilize are sustainable, carefully considering the potential social, societal and environmental impacts, risks and opportunities, including in the choice of partners and the way we build and use our solutions.
Transparency is essential for promoting understanding, trust, fairness and accountability, and should be translated in clarity, openness, and understandability of the processes involved in developing and using AI systems. We communicate openly and clearly using easy to understand language about the use of AI to all individuals impacted by AI systems (e.g., employees, candidates, and associates).
We clearly describe the capabilities, benefits, intended purpose, limitations, and the data used by such technologies, and provide simple explanations for any
automated outcomes and decisions made. We ensure that we allow the possibility to offer human intervention so that the decision criteria and output of the This starts by letting the end user know in advance when they interact directly with an AI system. We also prioritize the use of simpler, easier to explain models wherever this is possible without compromising their effectiveness.
AI system can be reasonably understood, challenged and validated by humans.
We will ensure that the data, AI systems, and the environments in which they are stored and operate are safe and secure, they will be technically robust. We ensure a quality management and risk management system is in place to govern the development and use of AI systems across their lifecycle.
Concretely, this means:
-
We ensure the AI technologies we deploy are safe to use and ensure those technologies do not go beyond the original purpose of their design.
-
Assure market-leading Information Security standards are used to secure data inputs and outputs of AI technologies.
-
Maintain oversight and control of AI technologies throughout their lifecycle.
-
Undertake best efforts to eliminate or reduce any potential risks associated with each respective AI technology for our employees, candidates, associates and clients.
-
Remain vigilant for areas of critical concern and stay up to date on the newest developments in the AI domain.
-
Create AI systems that are resilient and secure, ensuring a fallback plan in case of detected misuse or incidents.
-
We ensure that the development and use of AI technologies complies with all applicable laws, including but not limited to data protection and privacy laws, anti-discrimination laws, intellectual property laws, and legislation specific to AI.
-
We take steps to identify and remove actual or potential bias and discrimination from data sets and algorithms to ensure that the use of artificial intel- ligence technologies is fair to users and does not unjustly impact people based on sensitive charac- teristics including race, ethnicity, gender, nationality, sexual orientation, disability, and political or reli- gious belief.
-
We are accountable for the governance of our AI technologies and insurance of a continuous quality management of operationalized AI solutions.
-
We ensure that the use of Artificial Intelligence technologies respects the fundamental human rights of individuals as set out in the Universal Declaration of Human Rights and the EU Charter of Fundamental Rights, and we are guided by the UN Guiding Principles for Business and Human Rights and the UN Global Compact as well as the ILO Declaration on Fundamental Principles and Rights at Work.