Everyone is talking about artificial intelligence (AI) for a good reason. It’s already revolutionizing the world as we know it. And governments are well-poised to use AI to improve services and operations. But we can expect the AI conversation to shift into 2025.
Attend any conference of government leaders, and you will hear AI being discussed. Open any publication for civil servants, and you will see articles on AI. Ask a group of people what they are excited about, and half will tell you, “AI.” Ask that group what makes them nervous, and half will say to you, “AI.”
AI, especially generative AI technology like large language models, has opened our minds to new opportunities for machines to help us be more productive in ways we’d never considered.
Prediction: 2025 ushers in the new era of multimodel AI
While AI and generative AI (GenAI) can assist with mundane or repetitive tasks, they can also help simplify complex jobs, increase productivity and produce faster outcomes. Because GenAI is pervasive and easy to use, is saying “just use GenAI” the best way to solve a problem or answer a question?
Civil servants are discovering that GenAI can’t answer all questions or tackle all tasks. There are times when a large language model isn’t the right tool. Think about it this way: A hammer doesn't build a house. It takes a good plan, many tools and the right skills. Similarly, civil servants are learning that different AI technologies accomplish different tasks.
In 2025, governments will look to multimodal AI, combining different AI technologies to perform tasks with incredible speed and accuracy while minimizing cost. However, the looming concern of AI’s trustworthiness will continue to be a factor in adopting AI and Gen AI.
Prediction: Lagging GenAI adoption, but not for long
Recently, Coleman Parkes surveyed senior government leaders worldwide about their uptake of GenAI. The findings revealed three primary concerns.
- 78% worry about privacy.
- 77% are thinking security risks.
- 62% have governance at the top of their minds.
The insights offer a healthy dose of caution—and rightly so, as the implications of mishandling data or producing inaccurate output with AI can have dire consequences for the government, its employees or the people it serves. As governments seek to increase productivity with AI, responsibility will continue to govern its adoption.
In 2024, governments experimented with AI by conducting pilots, establishing usage policies, and analyzing new data and regulations. Next year, we’ll see varying degrees of each of these as governments adopt AI at different paces and with greater or lesser fervor. For the public sector, slow and steady will win the race.
Prediction: AI will be intertwined with data governance
Governments are recognizing that the quality of AI outcomes is mainly dependent on data quality. Civil servants have grappled with siloed data for decades, rife with duplication, errors and gaps. A renewed focus on data hygiene is necessary, and its importance will only increase.
In the coming year, we’ll see that clean data doesn’t just benefit AI; the data can benefit from AI. New tools are being developed to help clean data, such as one that resolves entities between different data sets, helps match records, and reduces duplication.
Prediction: The use of AI for investigations will gain momentum
From tax to law enforcement to social benefits to procurement and environmental compliance, investigations involve finding hidden patterns and trends, anomalies and relationships. While analytics and AI are instrumental in investigations – investigations can benefit more powerful AI technologies to catch bad actors. Criminals and fraudsters are constantly evolving and exploiting loopholes. They are using AI to target victims and even impersonate loved ones and trusted institutions. Fraud fighters and investigators must fight fire with fire.
Prediction: Limitations exist, but AI holds promise
“Do more with less” isn’t going away. Governments will continue to experience budget constraints, workforce shortages, spending taxpayer dollars wisely and challenges to provide services to citizens. By putting data, analytics and AI to work, governments can prevent fraud, waste and abuse or recover losses. Investing in technology can help address many of the limitations and obstacles. Governments can save several dollars in the long run for every dollar spent on software. Over time, the savings added up.
Onward to 2025
There is no question that AI's maturation will change the way we work and go down in history as a transformative technology. Now that the hype has settled, it’s time for us to intentionally and cautiously adopt AI solutions that will improve our productivity.
