Fewer than one in five AI ethics guidelines mention international labour standards
Most artificial intelligence (AI) ethics guidelines that deal with the world of work make no reference to international labour standards, according to a new working paper from the International Labour Organization (ILO).
The paper, published in October 2026, cites a review of 245 AI ethics guidelines. That review found more than 60% of them discuss AI in the context of work, but fewer than 20% mention the ILO or its standards.
The authors are Jordi Agustí-Panareda, a socio-legal researcher and ethics officer at the ILO in Geneva, and Daniel Samaan, a senior economist in the ILO's Research and Statistics Department in Geneva. They argue the omission is significant.
"The gap is particularly striking given that the governance of labour has long been shaped by international labour law with its well-developed normative system under the auspices of the ILO," they wrote.
Work risks acknowledged, standards left out
The underlying review, by Samaan and co-authors, is due to be published in the European Management Journal. It analysed guidelines issued by a wide range of actors using natural language processing and large language models.
It found that labour markets do not feature as a distinct topic, although many guidelines raise concerns about employment, working conditions or labour relations.
Despite this, the working paper notes, "few guidelines explicitly reference the International Labour Organization (ILO) or its International Labour Standards (ILS)."
"Much of the ethics discourse proceeds in isolation from the century-long normative architecture developed by the ILO," the authors wrote.
Collective rights the clearest blind spot
The paper compares five widely cited AI ethics principles – justice, beneficence, non-maleficence, autonomy and explicability – with the ethical foundations of the ILO's framework.
It finds important convergence on justice, beneficence and non-maleficence. It also finds that AI ethics focuses largely on individuals, while labour standards place greater weight on collective representation.
"For the right of collective bargaining, however, we do not see a clear ethical equivalent in existing AI ethics frameworks," the authors wrote.
They argue that the ILO's foundational commitments are internationally agreed and backed by supervisory mechanisms. As a result, those commitments "provide a level of clarity, legitimacy, and durability that most contemporary AI ethics frameworks typically do not yet possess."
What it means for responsible AI policies
The findings raise a practical question for HR leaders drafting or reviewing responsible AI policies: whether those policies account for workplace obligations such as non-discrimination, freedom of association and collective bargaining.
The authors argue that where AI informs decisions on hiring, performance evaluation or dismissal, employers need transparency about the data and reasoning involved, so that those decisions can be assessed for legal validity.
"AI ethics in the world of work should be developed not in isolation, but in closer dialogue with the ILO's normative framework, its tripartite governance, and its supervisory institutions," they concluded.
