This essay is republished from the John Curtin Research Centre's Tocsin Journal as the joint winner of the 2025 Henry Boote-David Cragg Young Activists Prize. Oscar Kaspi-Crutchett is currently serving as the Senior Research Organiser at Victorian Trades Hall Council.
A few years ago, a pandemic swept the earth. In the years after, rapid inflation drained workers’ savings. Old regimes fell, and new ones came to power. In many countries, radicals usurped moderate progressives.
Political violence, while sporadic, appeared on Australia’s streets. Technology, meanwhile, was transforming the pace, intensity and content of work. Revolutions in steel, chemicals and electricity allowed specialised crafts to be performed by a much broader pool of workers.
Stable, even prestigious, job roles became pliable and insecure. An economy-wide tendency set in towards great speed, regimentation, monotony, and surveillance: scientific management. Trust in politicians sunk. Even in the Labor base, an ‘acute sense of economic malaise’ fuelled murmurs that the interests of workers were ‘continuously sabotaged … by their parliamentary leaders.’
This was Australia in the early 1920s, as Henry Boote approached the heights of his significance as a labour intellectual.
In the decades following, the democratic world slouched further into ‘private affluence and public squalor’ – with its four attendant horsemen of inequality, populism, hopelessness, and violence. By the mid-century, these forces metastasized into a brutal crescendo: a war unparallelled in its destruction and carnage.
Australia finds itself, today, at a moment which echoes those years after the Great War. The stakes are, certainly, equally high: and although history is always particular to its context in a way that is ‘naïve to transcend,’ it does offer revelations that can help us chart the seas ahead.
Artificial intelligence (AI) is a machine that can analyse and generate text, produce images, manage inquiries, organise production, interpret scans and evaluate performance. It can administer care, dispense medications, educate children and counsel executives.
AI, therefore, is not simply a novel piece of capital equipment, but an entirely new factor of production. It will transform all existing productive inputs - above all: labour.
Much of the prevailing commentary about how AI will impact workers can be distilled into a single, underwhelming phrase: ‘it depends.’ This essay will answer what exactly it depends on: worker voice.
Worker voice means, specifically, the extent to which working Australians can take collective action that forces their concerns and preferences about AI-deployment to be recognised.
At the task-level, AI will shift workers from generating productive outputs and towards refining, proofing and editing the work of AI. Algorithmic task allocation will rapidly shift workers between stations according to instantaneous changes in market demand or operational need. Increased output will be achieved with fewer specialists. Complex tasks, usually undertaken by one well-trained in-house worker, will be simplified and distributed between several less-trained workers, outsourced labour, or AI itself.
At the firm-level, AI will become integrated into the functions of management: recruitment, coordination, discipline, instruction, termination.
The use of AI to justify and enforce managerial decisions necessarily reduces their contestability: ‘the computer says no.’ As algorithmic complexity increases, so too does this opacity. Already, roughly two-thirds of Australian employers use AI in hiring, firing and performance monitoring. Bias testing, risk management, and worker consultation remain sparse.
In many firms, algorithmic management will produce intense but highly repetitious work environments: ‘electronic sweatshops’ where work is quantifiable and strictly delineated, with minimal space for experimentation, spontaneity, variety, or discretion. The modern call centre is the best example of such a workplace, but many Australian factories, retail outlets and warehouses are coming to resemble it too.
In some industries, dynamic wage setting will see gig-economy style arrangements be introduced: where workers are used and discarded according to instant fluctuations in market activity – the Hayekian ideal realised.
Centralised, cybernetic management systems will coordinate sprawling but isolated workforces, ‘breaking down the social fabric that holds potential for worker power in the first place.'
AI will also drive a massive upscaling in workplace surveillance. This will allow employers to intensify workloads, identify opportunities for job destruction and eliminate operationally-induced gaps in production. The location data produced by wearable devices worn by warehouse workers not only enforce discipline, but also trains the robots that replace them.
In the worst cases, perpetual monitoring, job strain, and automation anxiety will surge workloads and multiply hazards.
Industrial-era unionbusting may also reappear: censorship, blacklisting, espionage. Amazon uses AI to forecast if warehouses will unionise. Deliveroo’s algorithm punishes riders who join union protests. Woolworths deployed an AI-driven performance monitoring system to target union delegates. Rushed automation efforts will endanger workers’ lives. Last year, a 27-year-old warehouse worker was crushed by an automated robot at a Melbourne distribution centre.
The most automated Amazon warehouses in the US have the most worker injuries.
At the economy-wide level, we arrive at the most significant dimension of this issue: how AI will adjust the quality of work, wages and national income shares that workers can command with their skillsets.
Even workers in stable jobs will face immense downward pressure on wages as they compete not only against machines but novices who can use AI to rapidly ascend the skill curve. AI will make many jobs simpler, and a very small number significantly more complex. Job-losses will occur, but overall, AI will likely increase demand for lower-paid, insecure work. The destruction of each full-time role in a factory or office will be paired with the creation of a new, more precarious, position elsewhere.
Over time, jobs and incomes will polarise – and the general configuration of society may regress to its premodern form.
The effects of AI on collective bargaining could accelerate this process. Workers’ ability to bargain for a better life is determined by several factors:
- The asymmetry of information between workers and employers about each party’s relative power;
- The degree to which the work environment allows for workers to develop bonds that can form a basis for solidarity, and
- Whether workers possess productive knowledge that can be strategically leveraged during bargaining.
AI augments all of these. As employers automate key duties currently performed by humans, they neuter the disruptive potential of industrial action:
‘I’m siphoning off his knowledge … he is replaceable now.’
Automation, therefore, will be initially concentrated at chokepoints where workers retain strategic advantage. Efficiency considerations are secondary. Even where businesses are inclined towards restraint, competitive pressure will drive a race to the bottom.
There are no demons or devils here: only the forces of the market.
The role for unions, therefore, is very clear.
While regulators have an incentive to prevent harm, they are not proximal to AI deployment and can only respond to it after-the-fact. Employers, meanwhile, are proximal to deployment, but structurally incapable of restraining it for collective ends.
Only organised workers have the combination of proximity and incentive necessary for steering AI towards human enrichment.
The collective agreements of the future will include clauses which give workers:
- The right to contest status-altering automated decisions,
- To receive a share of profits made by the sale of their data,
- To order union inspections of task allocation systems,
- Veto covert surveillance,
- Commission bias testing,
- Enforce directorial liability for AI-driven discrimination, and
- Issue stop-work orders when AI deployments imperil public safety.
Both intellectually and tactically speaking, Australian workers are well-positioned to face our century’s challenges. This is demonstrated by their world-leading rates of AI scepticism and concern.
Many elites decry this lack of enthusiasm for AI as an embarrassing marker of national parochialism – urging workers to ‘get educated,’ ‘upskill,’ and trust the technology despite its lack of social license. In fact, widespread trepidation is not only entirely appropriate, but a shining demonstration of the ever-sound and ever-sober instincts of our national community.
Removed from the tumult of the old world, Australian workers and their unions have always been ‘resistant to both dreams and doctrines,’ invariably drawn to ‘being practical.’ On AI, that means caution and concern. Laissez faire is radicalism.
The long memory of the labour movement affords it a certain maturity not available to professional futurists or prophets of doom.
We understand that innovations like AI always carry immense destructive and generative potential. The process which translates their emergence into broad prosperity is not automatic or guaranteed. It demands political courage and tireless struggle. We have done this before.
For the union movement, strategic myopia, timidity, and bureaucratic inertia are all hazards - but so too is overestimating the modern workers’ zeal for upheaval and reconstruction. As one prominent labourite in Boote’s day warned, ‘hitch your wagon to a star,’ but remember, ‘the people want Labor to do something for them today.’ History demands a visionary turn. We must take it with clear eyes and steady hands.
Unmanaged, the AI rollout presents incendiary risks for the future of work. Steering it towards fairness is the greatest challenge of our time: a fight that will either restore the movement to its lost zenith or deliver the final blow in its annihilation. History watches, as do our descendants. I conclude, here, with the words of Boote himself:
“Everything is in the melting pot. Labor too. It must show that it can survive the ordeal, and emerge imbued with a new vitality. It must show that it is swift to seize upon the psychology of the times and turn it to the purposes of progress.
If it cannot do that … then assuredly when the melting pot is emptied you will have to look for Labor in the dross that remains.”
References
Primary Sources
Australian Labor Party. Official Report of Proceedings of the Ninth Commonwealth Conference of the Australian Labor Party. Brisbane, October 1921. Commonwealth Parliamentary Library: 1340111.
Boote, Henry. Australian Worker, ‘What of the Brisbane Conference?’ 20 October 1921, p. 3.
The Sun, ‘Industrial Labour, 2 September 1921, p. 8
Tribunal Order, Bologna Labour Court, 27 November 2020, Case No. 2949/2019, Italian General Confederation of Trade Unions v Deliveroo Italia (It.), available at https://www.bollettinoadapt.it/wp-content/uploads/2021/01/OrdinanzaBologna.pdf
Secondary Authorities
Books
Galbraith, K. The Affluent Society. Boston Massachusetts: Houghton Mifflin, 1958.
Gollan, R. The Coal Miners of New South Wales: 1860-1960. Melbourne: Melbourne University Press, 1963.
Perelman, M. Class Warfare in the Information Age. New York: St. Martin's Press, 1998.
Book Chapters
McQueen, H. “Shoot the Bolshevik! Hang the Profiteer.” In Essays in the Political Economy of Australian Capitalism, edited by E.L Buckley and K Wheelright, K. Sydney: Australia and New Zealand Book Co., (1978), pp. 185-286.
West, D. “A New Social Contract” in The Future of Work: Robots, AI and Automation. Brookings Institution Press, Washington D.C, (2018) pp. 89-108.
Reports
Delfanti, A, Radovac, L and Walker, T. UNI Global Union. “THE AMAZON PANOPTICON: A Guide for Workers, Organisers and Policymakers,” 2021.
IndustriAll: European Trade Union. “Artificial Intelligence: Humans must stay in command.” Policy Brief.
Juego B et. al., “Algorithms By and For the Workers: Towards A Fair, Democratic and Humane Digitalisation of the Workplace.” Brussels: Foundation For European Progressive Studies, Policy Study, 2024.
Staab, P and Prediger, LG. “Digitalisation and Polarisation: A literature review on the effects of digital change on social structures and companies.” Dusseldorf: Research Institute for Social Development (2019).
Trades Union Congress. United Kingdom (2020). “Technology Managing People: The Worker Experience.” London: Trades Union Congress.
United Workers Union. Disciplinary Power in Australian Grocery Warehouses: Engineered standards, algorithmic management, and the veneer of technological objectivity. Submission to the Inquiry into the Digital Transformation of Workplaces. June 2024. Available at: https://unitedworkers.org.au/wpcontent/uploads/2024/07/Disciplinary-Power-in-Aust-Grocery-WarehousesUWU.pdf.
Journal Articles
Atkinson, J and Collins, P. “Worker voice and algorithmic management in post-Brexit Britain.” Transfer: European Review of Labour and Research. Vol. 29, No. 1, (2023) pp. 37-52.
Autor, D and Dorn, D. “The Growth of low-skill service jobs and the polarisation of the US labor market.” American Economic Review, Vol. 103, No. 5, (2013), pp. 1553-1597.
Bertossa, D and Voss, E. "Collective Bargaining and Digitalization: A Global Survey of Union Use of Collective Bargaining to Increase Worker Control over Digitalization," New England Journal of Public Policy. Vol. 34, No.1, (2022) pp. 1-36.
Christiaens, T. “Nationalise AI!” AI & Society: Knowledge, Culture and Communications. Volume 40, (2024), pp. 1147-49.
Cronin, JE. “Labor Insurgency and Class Formation: Comparative Perspectives on the Crisis of 1917-1920 in Europe.” Social Science History, Winter, 1980, Vol. 4, No. 1. (1980), pp. 125-152.
De Stefano V and Doellgast V. “Introduction to the Transfer special issue. Regulating AI at work: labour relations, automation, and algorithmic management.” Transfer: European Review of Labour and Research. Vol. 29, No. 1, (2023), pp. 9-20.
De Stefano, V and Taes, S. “Algorithmic management and collective bargaining,” Transfer: European Review of Labour and Research, Vol. 29, no. 1, (2023), pp. 21-36.
Dyrenfurth, N, and Quartly, M. "Fat Man vs. 'The People': Labour Intellectuals and the Making of Oppositional Identities, 1890-1901." Labour History, no. 92 (2007), pp. 31-56.
Fuchs, Christian. "Political Economy and Surveillance Theory." Critical Sociology, Volume 39, no. 5, (2013), pp. 671- 687.
Jayakumar, S and Goh, E. “The Future of Work: New Underclass, Dystopian Reality?” S. Rajaratnam School of International Studies, Vol. 203 (2017). RSIS Commentaries, Singapore: pp. 1-5.
Kinowska H and Sienkiewicz LJ. “Influence of algorithmic management practices on workplace well-being – evidence from European organisations.” Information Technology & People. Vol. 36, No. 8, (2023), pp. 21-42.
Klengel E and Wenckebach J. “Artificial intelligence, work, power imbalance and democracy – why codetermination is essential.” Italian Labour Law e-Journal. Vol. 14, No. 2, (2021), pp. 157-171.
Manokha, I. “The Implications of Digital Employee Monitoring and People Analytics for Power Relations in the Workplace.” Surveillance & Society, Volume 18, no. 4 (2020), pp. 540-554.
Mendonca P and Kougiannou N. “Disconnecting labour: The impact of intraplatform algorithmic changes on the labour process and workers’ capacity to organise collectively.” New Technology, Work and Employment, Vol. 38, (2023), pp. 1-20.
O'Farrell, PJ. “The Russian Revolution and the Labour Movements of Australia and New Zealand, 1917–1922.” International Review of Social History, Volume 8, No. 2 (1963), pp. 177-197.
Scalmer, S. “Being Practical in Early and Contemporary Labor Politics: A Labourist Critique.” Australian Journal of Politics & History, Vol. 43 (2008), pp. 301-311.
Schaupp, S. “Algorithmic Integration and Precarious (Dis)Obedience: On the Co-Constitution of Migration Regime and Workplace Regime in Digitalised Manufacturing and Logistics.” Work, Employment and Society, Vol. 36, No. 2 (2022), pp. 310-327.
Schaupp, S. “Cybernetic Proletarianization: Spirals of devaluation and conflict in digitalized production.” Capital & Class. Vol. 46, No. 1, (2022), pp. 11-31.
Skinner, Q. “Meaning and Understanding in the History of Ideas." History and Theory, Volume 8, No. 1 (1969), pp. 1-59.
Public Opinion Surveys
“Australian Public Attitudes Toward AI Risks and Governance.” The University of Queensland; GovTech Review. 3 September 2020. “Research explores Australian public attitudes towards AI,” Available at https://www.govtechreview.com.au/content/gov-tech/news/research-exploresaustralian-public-attitudes-towards-ai-1420472666.
Fifth Quadrant, 2024. “The Australian Responsible AI Index.” 5 September 2024. Available at: https://www.fifthquadrant.com.au/responsible-aiindex-2024
Gillespie, N., Lockey, S., & Curtis, C. “Trust in Artificial Intelligence: A Five Country Study.” The University of Queensland and KPMG Australia, (2021).
Gillespie, N., Lockey, S., Ward, T., Macdade, A., & Hassed, G. “Trust, attitudes and use of artificial intelligence: A global study 2025.” The University of Melbourne and KPMG (2025).
Noetel, M., Saeri, A.K., Graham, J.G., Slattery, P. Survey Assessing Risks from Artificial Intelligence (SARA), “Australian Public Attitudes Toward AI Risks and Governance.” The University of Queensland, (2025).
Other
Calacci, Daniel. “Organizing in the End of Employment: Information Sharing, Data Stewardship, and Digital Workerism,” in CHIWORK ’22: The Symposium on Human-Computer Interaction for Work, June 08-09.
Massachusetts: Massachusetts Institute of Technology Media Lab, 2022.
Ciuriak, D. “Machine Knowledge Capital and the Future of Work.” Available at SSRN: https://ssrn.com/abstract=4753468. 2024.
Employee Matters. “Almost half of all Aussie workplaces admit to using AI to spy on their employees.” 24 October 2024. Available at: https://www.employeematters.com.au/yes-your-boss-is-spying-on-you-do-theyneed-their-own-watchdog/
WorkSafe Victoria. “Worker dies after being crushed by machinery” 10 September 2024. Available at: https://www.worksafe.vic.gov.au/news/2024-09/worker-dies-after-being-crushed-machinery