The Rise of AI in Public Services — Efficiency at What Cost?
As artificial intelligence (AI) and machine learning (ML) make strides in transforming public services, their role in customer service, especially through chatbots and digital assistants, raises critical questions about trust and reliability.
Although these technologies promise efficiency, accuracy, and personalised interactions, they are not yet fully equipped to handle complex cases or sensitive information independently.
The Norwegian government’s digitalisation strategy for 2024 to 2030 underscores that while AI can unlock substantial benefits for public services, its deployment must be carefully managed to uphold ethical standards and safeguard public trust. Particularly in customer-facing roles, the strategy emphasises the necessity for a human intermediary to ensure quality control and validate information before it reaches the user, especially when legal and informational complexities are involved.
This approach not only helps foster trust but also aligns AI deployment with the high standards required in the public sector. By integrating AI responsibly, with human oversight, we can support technological advancement without compromising accuracy, security, or user confidence.
Elders vs. Digital Natives: Who Navigates Bureaucracy Better?
Even though the focus of the digitalisation strategy these past two decades has been on increasing the digital literacy of the elderly, it has left out the fact that research shows that the elderly with low digital skills navigate bureaucracy far better than the younger generation. Being a digital native or literate in technology doesn’t necessarily translate into the ability to navigate complex government systems like welfare. Even though the younger generation might be tech-savvy, they are missing the institutional knowledge needed to understand bureaucratic processes, legal jargon, and specific rights they have within these systems.
Digital Naïvity and Its Consequences
The younger generation is becoming increasingly comfortable with AI tools, such as language models, seamlessly integrating them into their daily lives — often without even realising they are using them. Research shows that repeated positive interaction with personalised and intelligent services can gradually create a false sense of security, as they act as a support system compensating for users’ lack of competence and critical thinking. This leads to what I call digital naïvity.
Digital naïvity refers to a state in which individuals, particularly those who are comfortable using digital tools, have an over-reliance on technology without fully understanding its limitations or the broader context in which it operates. It reflects a false sense of competence — people may assume that because they are proficient in everyday digital skills (like using apps, social media, or basic AI tools), they are also capable of navigating more complex systems, such as government services, legal frameworks, or security-critical environments.
Trust in AI and Public Services: A Higher Bar
When everyday services, such as Siri, Alexa, Instagram, or Spotify make mistakes, users usually dismiss them with a shrug. However, in critical, high-security domains like public services, trust in these technologies remains a significant challenge, particularly where the consequences of failure are amplified.
We set the bar significantly higher regarding trust when interacting with machines compared to humans. We are much more forgiving towards humans because we expect them to make mistakes. Machines, on the other hand, are expected to perform consistently and flawlessly.
When dealing with the critical nature of public services, unlike consumer-focused sectors, these services often involve legal rights, welfare decisions, or healthcare access — areas where there’s little room for error.
AI systems are held to higher standards, partly due to the perception that machines should be “perfect” because they are built by humans, and thus their failure feels less acceptable. This psychological aspect creates a unique trust barrier — when AI gets it wrong, the public might interpret the mistake as a systemic flaw rather than an anomaly, setting the implementation of this technology back years.
“Some decisions are too complex and too big to be left to business leaders alone.”
-Jens Stoltenberg, former Prime Minister of Norway and Secretary General of NATO
Reverse-Engineering Strategies and Complex Decision-Making
Stoltenberg’s observation about the politicisation of energy trade agreements mirrors the challenges public sector executives face when deciding on the use of AI and ML in personal assistants. These technologies hold immense potential, but their adoption is influenced by ethical debates, public trust, and political agendas. In areas like customer service and chatbot-driven public services, the use of AI must be carefully navigated to balance innovation with citizens’ rights and trust — an inherently political endeavor.
As the complexity of technology grows, public sector executives often struggle to fully understand its nuances, which leads to reverse-engineering their strategies — starting with technological solutions and working backward to address their needs. Instead, I argue that the growing complexity of technology demands critical thinking and cross-disciplinary collaboration rather than relying solely on technology to solve problems. The consequences of these decisions are not always apparent to executives pursuing short-term economic gains, as they involve societal and psychological impacts, particularly regarding public trust.
Such decisions require collaboration between technical experts, legal advisors, regulatory bodies, and ethical committees. A cross-disciplinary approach is important to make sure business goals are aligned not just with profitability but with sustainability, legal certainty, and ethical considerations.
Why Human Quality Assessment Still Matters
Don’t get me wrong, I’m a technology optimist. I love how technology can make our lives easier and our work more efficient. But we need to recognize that implementing such technology fundamentally depends on trust. Therefore, using technology to empower knowledge experts, such as those in contact centers, offers an incredible opportunity. By helping these experts quickly find and convey accurate information to users, technology can act as a filter between complex systems and the end user. This approach not only makes the lives of those who advise others easier but also helps reduce the risk of errors, ultimately leading to better outcomes for everyone involved.
The role of humans as quality assessors and as facilitators between technology and users is a key component of HCI that has gained renewed importance. So, in many ways, HCI is not just back — it’s taking on an even more crucial role in ensuring that as technology advances, it does so in ways that are truly beneficial and equitable for everyone.
The Balance Between Innovation and Trust
While AI and automation promise efficiency, maintaining human oversight is essential for trust and ensuring equitable outcomes. Public sector leaders must adopt thoughtful, human-centered approaches to technology integration to truly benefit all citizens.
