27 September 2018
The OECD is moving ahead with its work on Artificial Intelligence (AI) towards developing a Council Recommendation in the course of 2019 and creating an AI observatory as a pillar of the second phase of the Going Digital project.
On 24-25 September, the first meeting of the AI expert Group at the OECD (AIGO) was held in Paris. It will be followed by other meetings in the next months. The AIGO has multi-stakeholder membership and brings together experts who have been nominated by OECD delegations and advisory committees (including the TUAC) as well as a handful of invited experts. TUAC was represented through its Secretariat, Anna Byhovskaya and by UNI Global Union, Christina Colclough. The group is asked to contribute to the scope and content of the OECD principles on AI – which are expected to range from safeguarding core democratic and societal values, to operational parameters for AI systems, to guidelines for policy frameworks.
At both, the national and regional level (including the EU level) such multi-stakeholder discussions are currently underway. The implications of an increasing and wider use of AI, combined with other technologies will be far reaching on jobs and on workers – from changing work patterns, to data control, to the replacement of tasks. Yet, many opportunities and challenges of AI and its scope can still not be fully grasped as its development and spread take place at an unprecedented speed.
For TUAC, building in safeguards into design, development and the use of AI will be crucial, not least on employment aspects. Many policy proposals start and end with education and training when it comes to societal impacts. Collective bargaining or just transition funds are not on the radar in some of these debates, yet.
Public policy needs to look into the economic, social including labour market, ethical and legal aspects of AI as several risks arise. At the same time, it is important to keep the balance between applying and revising existing standards and regulatory frameworks. The TUAC had published a first set of priorities in November 2017 (https://tuac.org/news/shaping-introduction-ai-benefit/ ) and will flash these out further: