A prescient Policy Brief from the OECD in 2001 examined ‘The Hidden Threat to E-Government.’ According to the authors: ‘Most governments experience problems when implementing large IT projects. Budgets are exceeded, deadlines are over-run and often the quality of the new system is far below the standard agreed when the project was undertaken.’ 

More than two decades later, the sobering message still resonates. From the Phoenix Payroll debacle to ongoing struggles with IT Modernization (backend data centres and front-end legacy benefit systems), and the now-infamous ArriveCan app, it would seem that the foundations of digital government transformation remain all too shaky. 

Even the former federal government CIO agrees. Questioned last year about the Auditor General’s harsh review of digital infrastructure readiness, Catherine Luelo conceded to Parliamentarians (as reported by IT World Canada): ‘I think [the Auditor General’s report] is a good and accurate reflection of where we find ourselves in terms of our current state of technology. We’ve not advanced in the last 13 years, and I don’t think that is a win for Canadians.’ 

So why the ongoing penchant for systemic failures? There are arguably three sets of inter-related factors: first, a lack of consistent political leadership and support for digital renewal; secondly, an antiquated governance architecture overly-centralized within the Treasury Board Secretariat (TBS); and thirdly; under-investment into internal workforce and organizational capacities.  

With the rise of Artificial Intelligence (AI), the implications of these shortcomings are more profound than ever. As just one example, CBSA’s mishandling of the ArriveCan app (a shared failure with other federal entities) portends wider problems ahead for its ambitious Border Modernization initiative that will feature significant AI elements. The absence of a public oversight body specific to CBSA (a new Review Commission has been stalled in Parliament for years) reinforces insularity and further limits public trust. 

In some facets of AI governance, the Government of Canada deserves credit: the Algorithmic Impact Assessment framework and the recent internal guidance on the usage of Generative AI are cases in point. A useful report by the federal AI Advisory Board further illuminates the economic and societal challenges ahead. Yet unless the federal government can get its act together and better manage itself digitally, the accelerating pace of change inherent in AI will invariably prove overwhelming (with significant risks and unintended consequences).  

Let’s return to the three sets of digital shortcomings plaguing the Government of Canada (leadership, architecture, and internal capacity-building). The ongoing absence of a federal Minister for Digital Government coupled with the vacancy of a permanent CIO – as well as the notable void of a senior level position devoted to AI, all speak volumes to the dearth of top-level leadership at present. 

By contrast, fueled by a Presidential Executive Order, the US federal government has convened an internal White House AI Council compromised of senior officials from several digitally-minded agencies. Among other strategic initiatives, the Council is overseeing an ‘AI Talent Surge’ to accelerate the hiring of AI professionals and data scientists, while a separate AI Task Force has been struck to exploit the promise of AI innovation in health care. 

With respect to a digital architecture shackled by a TBS culture of financial austerity and managerial control, it’s long overdue for the Canadian Government to take a page from the Australian playbook and create new entities (such as the Digital Transformation Agency) with a uniquely digital mindset and focus. In health care – a centrepiece of promising AI innovation, as the US federal government has recognized, a new and federated equivalent to Australia’s National Digital Health Agency is sorely needed. 

In terms of weak capacities, the ArriveCan app is merely the latest example of a tendency to overspend externally and underinvest internally. As Carleton University’s Amanda Clarke told Canadian lawmakers in 2023, the reliance on management consulting firms within the public service ‘betrays’ the principles of responsible public administration. Even external partners agree: the Boston Consulting Group has made building AI capacity within government a centrepiece of its messaging and advice.  

In sum, before deepening AI experimentation, as it should be doing with a healthy mix of ambition and skepticism, the Government of Canada must double down on addressing the glaringly persistent and longstanding threats to public sector digitization, namely: insufficient leadership, antiquated governance, and ailing internal capacities.