Gartner predicts that by 2027, less than 25% of government agencies will have generative AI services for citizens

Factors that affect adoption include fear of public failure and social disapproval

Gartner predicts that by 2027, the percentage of government agencies with generative artificial intelligence (generative AI) services for citizens will be less than 25%. Factors slowing government’s external use of this technology for citizen services include fear of failure in public and a lack of trust in government’s use of this technology.

Like all industries, governments have spent the past 15 months exploring the opportunities and risks presented by the emergence of generative AI. Gartner’s annual global survey of more than 2,400 CIOs and technology executives found that 25% of governments have deployed or plan to deploy generative AI in the next 12 months; another 25% plan to deploy generative AI in the next 24 months. In the early days, governments focused on establishing an initial governance framework to support trials and small-scale adoption.

“While governments have benefited from using more mature AI technologies for years, they are now slowing down their large-scale adoption of generative AI due to risks and uncertainties, particularly the lack of traditional controls to reduce drift and illusions,” said Dean Lacheca , research vice president at Gartner. “Also, public acceptance of generative AI for citizen services may be affected if governments lack empathy in their services and fail to meet society’s expectations.”

Adjust the speed of adoption based on risk tolerance

To address this, Gartner recommends that governments continue to improve internal aspects of citizen services by actively deploying generative AI solutions.

“To prevent early AI missteps from dampening societal acceptance of the technology for government services, agencies should decide how fast to adopt generative AI based on their risk tolerance,” Lacheca said. “This means that governments will adopt AI in the background faster than they use it to directly serve citizens.”

Gartner recommends that government agencies can accelerate the adoption of generative AI by focusing on use cases that primarily affect internal resources, preventing perceived risks associated with citizen services, and accumulating knowledge and skills related to the technology. In addition, governments should establish a transparent AI governance and assurance framework for internally developed and purchased AI capabilities to build trust and reduce related risks.

“ These frameworks need to specifically address the risks associated with citizen service use cases, such as inaccurate or misleading results, data privacy, secure conversations, etc. To do this, ensure that the governance process is specifically addressing each risk, both before and after the initial implementation,” Lacheca said.

In addition, government agencies should adopt human-centered design practices centered on empathy when designing and using AI solutions for citizens or workers. This can ensure that society’s expectations are met when determining how and when these solutions should be used from the perspective of citizens.

Gartner clients can learn more in "Predicts 2024: Government CIOs must prepare for an AI-enabled future."