“Human intelligence must dictate AI at work” – Séamus Dooley

Labour history conference to hear call for
ILO Convention on digital transformation

The need for national governments, European and international organisations to develop a rights based approach to the application of artificial intelligence in the workplace has been stressed by Séamus Dooley, Assistant General Secretary of the National Union of Journalists.

NUJ Assistant General Secretary Seamus Dooley

Mr Dooley will this evening deliver the annual Constance Markievicz Memorial lecture at Liberty Hall, Dublin, as part of the Irish Labour History Society 47th Anniversary conference, “Visons of Labour and Class in Ireland.

Speaking in advance of the address Séamus Dooley said the International Labour Organisation should develop an international convention on the use of transformational digital technologies predicated on the right of workers to have a say in changes in their working lives. Humans, not technologies, should dictate change.

Referring to the report of the High-Level Group on Collective Bargaining published last October Mr Dooley said the issue of collective representation of workers would become more urgent with the advent of AI. Workplace change brought about by AI should be subject to collective agreements.

“The right to be recognised and to be consulted on workplace change is of fundamental importance. It is unconscionable that companies would be allowed introduce technological changes such as Generative AI without collective agreements.”

The evolution of Generative Artificial Intelligence (AI) poses an existential threat to the future of journalism and to society.

The rapidly advancing technologies which seek to displace human thoughts and actions with machines may initially have been perceived as having a sort of sci-fi novelty quality but as the seismic changes enabled by Artificial Intelligence become apparent, the global trade union movement is waking up to the reality that this latest wave of new technology poses an existential threat beyond our wildest imaginings.

The ILO study, Generative AI and jobs: a global analysis of potential effects on job quality and quality published last month[1] suggested that AI is more likely to augment than destroy jobs by automating some tasks but finds, unsurprisingly, that AI will have a major impact on the quality of work. The report suggests that most jobs and industries are only partly exposed to automation and are more likely to be complemented rather than substituted by the latest waves of Generative AI. I would not be so complacent.

Of particular concern is the finding that the potential effects of Generative AI are likely to be different significantly for men and women, with more than twice the share of female employment potentially affected by automation. This is due to the over-representation of women in clerical work, especially in high and middle-income countries.

Clerical jobs have traditionally been an important source of employment as countries develop economically. One result of Generative AI could be that certain clerical jobs may never emerge in lower-income countries. This prediction has profound implications and there must be a global commitment to the provision of training to ensure that women workers, in particular, are not further marginalised by digital exclusion.

Unions must be cognisant of the opportunities AI presents for workers when used ethically but the chorus of caution from journalists, creators, and artists alike must not be ignored.

The reinforcement of bias through the use of AI in recruitment and selection procedures is a reality which cannot be dismissed. The potential for bias and prejudice in shortlisting, the absence of human involvement in shortlisting and selecting candidates should be a major concern.

Global change requires a global response. The ILO must be to the fore in ensuring that workers voices are heard, and that digital transformation does not re-enforce exclusion, not just in industrial sectors but on a geo-political basis. As the report noted, there is a danger that only a few, well-prepared countries and marker participants will benefit from new technologies.

I believe there would be real value in an ILO Convention on Digital Transformation, although I recognise that the speed at which AI is already transforming the world of work exceeds the speed at which global institutions tend to respond.

The decisions to incorporate AI are taken by humans. It is imperative that it is humans who guide and prescribe how the transition process develops and that it is human values which inform policy decisions. Governments, European and international agencies must not cede control to corporations.

There can be no ambiguity about what is generated by AI and what is produced by humans – this must become the legal norm.

Just as additives are featured on food labels setting out all ingredients, it is important that AI generated copy is clearly marked. Historians and scholars will be hard pressed to determine the veracity of contemporary records and will have even more reason to challenge the notion of journalism – on whatever platform – as the first draft of history.

Algorithms, the use of intelligence to guide readers towards ideas which re-enforce beliefs, preferences and prejudices, pose a threat to intellectual discourse.

Speaking at the World Congress of the International Federation of Journalists in Dublin in 2013, President Michael D Higgins warned of the dangers inherent in the editorial power being amassed by search engines.

He told delegates: “We can see the editorial power being granted to search engines – it is not a huge leap to suggest that in the future, these adaptive and hugely useful technologies will come to exercise an increasingly powerful role in how people will access media. Similarly, it is easy to see how this globalisation of content might allow popular commercial material to become the exclusive preserve of large multinational content providers – vertically integrated media companies that might come to control the commercial middle ground and with it access to the platforms people use to view content.”

These concerns have already been shown to be fully justified.