Judith Collins, our Minister for Digitising Government, says “we stand at the cusp of a digital revolution that has the power to transform the way our government serves New Zealanders.” According to the Minister this will mean agencies moving at pace to do this by “leveraging AI and other data-driven technologies.”[1] Already 65% of agencies report they are using AI and 85% plan to.[2]
What will this revolution mean for the jobs of people working in public and community services and how can we shape digitised public services so that they operate for the public good and in the public’s interests?
The introduction and use of AI is part of the ongoing digitisation of work. Technology shocks (the rapid introduction of new technologies) are always disruptive for workers and labour markets, and there are always winners and losers. The PSA has been around as a union since 1913, so we have seen and influenced the roll-out of generations of new technologies. We have a proud history of organising to ensure PSA members experience a just transition through technology change. Our focus is on ensuring workers in public and community services have a major influence on the introduction and use of AI in public and community services; and that this creates outcomes that maximise the benefits and prevent negative impacts of AI for workers and the public.
Generative AI and forms of automated decision making can offer many benefits for public and community services. Work can be done more efficiently, service design and delivery improved, and regulatory and compliance work can be speeded up. For example, councils and agencies can use automated decision making to grant compliance certificates or allocate grants. Hospitals can use it to prioritise patients for treatment and schedule appointments and correctional facilities can use it to classify people in prisons and allocate them to facilities and programmes. Chatbots can answer basic queries, and large language processing modules can help lawyers prepare documentation or policy people summarise public submissions. Developers can use AI to review code. All of this can free up workers to do work that adds more value.
But the downsides are also very real. Public agencies are rightly held to a higher standard in respect of trust and confidence and so safeguards are needed to ensure protection of personal, client and worker data. Those safeguards are not yet in place. Generative AI can produce inaccurate and incomplete outputs and has the potential to perpetuate bias and mis- or disinformation, and AI systems can be complex to understand, causing issues where understanding decision-making processes is important. Humans must still be the final decision makers and accountable for those decisions.
There are also important questions about data sovereignty to consider including in our New Zealand context not using mātauranga Māori in AI, and the environmental impact of AI tools, all of which require significant use of electricity to operate. Ownership, and so control of the tools and who benefits is important. In the US where AI is in greater use, ownership and so control is firmly with the private sector.
In a work and employment context, there are risks to workers’ privacy and with the protection of your personal (including biometric) information, risks to health and safety through increasing work pace, and risks to jobs as tasks are automated. New tech in public service workplaces should be used to empower people and enhance their work, not surveille them.
Other countries have been legislating to create guardrails for the use of AI in public services and at work. We can’t rely on that happening here. In late June, Cabinet made a decision to take a “light touch” approach and not regulate but rather to rely on our existing human rights, privacy and consumer protection laws – none of which were developed with AI in mind. This “agile approach” will rely instead on “voluntary guidance, industry codes, technical standards, and audit requirements”.[3] At the time of writing, Cabinet was due to decide a “principles-based AI framework to support responsible and trustworthy AI adoption and innovation in Government”, with no consultation with public and community workers or the public.
Where use of AI tools is being considered, those who will be using them need to have a say on purpose, design and implementation. In the absence of the government setting standards and safeguards through legislation, or other transparent democratic process, the PSA and other public and community service unions have a role in advocating for robust standards and safeguards at the agency and industry level. Together we can create outcomes that maximise the benefits and minimise negative impacts of AI for workers and the public.
[1] https://www.beehive.govt.nz/speech/speech-sixth-annual-new-zealand-government-data-summit
[2] Nov 2023, survey by Government Chief Digital Office.
[3] Approach to work on Artificial Intelligence, 26 June 2024, ECO-24-MIN-0119.
