More than 75% of IT leaders agree that generative artificial intelligence (AI) will significantly transform their organizations, but “AI readiness” in the public sector is shockingly low. By their nature, many public sector organizations lack the agility and financial resources to quickly adopt and integrate AI technologies into their operations, and many of their operational systems rely on aging technology. In addition, some agencies have ethical concerns about AI involving fairness, transparency, privacy, and human rights, adding further delays in their AI adoption.

Indeed, implementing AI can be risky if it’s not done right. But when public sector organizations consider AI with a thoughtful and deliberate approach, they are more likely to find success with improved efficiency and cost savings. With strategic implementation, AI also creates greater opportunity for citizen engagement via digital platforms, creating a lean service delivery without compromising quality.

AI Readiness in the Public Sector

AI readiness requires careful planning, strategic alignment, and a comprehensive understanding of an organization’s landscape. Public sector entities must not only consider the technological aspects of AI, but also how it will be deployed, adopted, and used in the workforce and in public-facing applications involving their community. These AI strategies must enable organizations to rapidly pivot to adapt to evolving social, technological, economic, and environmental trends.

As the adoption of AI and related applications accelerate change and disrupt the public sector, successful organizations are transitioning from outdated contingency planning models to a philosophy based on “readiness for constant change.” Once AI is implemented, they must be prepared for continuous change. But before starting AI integration, organizations should conduct a comprehensive AI readiness assessment to help ensure seamless transition in the fluid world of AI.

AI Readiness Assessment: A Checklist for Success

A strategic AI readiness assessment considers five broad areas: opportunity discovery; data management; IT environment and security; risk, privacy, and governance; and adoption.

Opportunity Discovery

Organizations identify areas where AI can add value and develop a business case to justify investment on AI initiatives. Public sector entities need to identify and define key strategic goals and objectives for their organizations. Potential participants may include employees, residents and end users, officials, contractors, and regulators. Organizations may also research best practices about how other similar public sector entities have deployed AI.

Data Management

Organizations identify key data stakeholders and identify, profile, and catalog existing data assets. The process varies among public sector entities. States and large cites, for example, may have more mature and complex data sets with vast amounts of information, while a small town may have smaller, simpler datasets. In this phase, the organization analyzes and classifies datasets based on their use, availability, sensitivity, and utility.

IT Environment and Security

Public sector entities need to evaluate their hardware, software, licenses, and security controls to ensure a safe environment for AI usage. They should also look at data storage, sharing and retention policies, and the unique security requirements for AI. In addition, organizations will need to review user accounts and access considerations.

Risk, Privacy, and Governance

With a deluge of sensitive information at their fingertips, organizations need to implement a risk management framework, data privacy controls, and governance processes to ensure safe data usage and compliance with regulatory frameworks. They need to certify that content created by AI is subject to governance processes, including content review and validation prior to releasing externally. They should also analyze organizational risk associated with the use of AI and measure the risk profile against key AI risk benchmarks.1 Because they deal with an extensive amount of personal information, public sector entities also need to consider data privacy and protection regulations and best practices.

Adoption

For successful adoption and integration of AI, public sector entities need to leverage effective change management practices to enhance data and AI literacy, create shared understanding, build trust, and drive cultural alignment to democratize the use of data and AI within the workforce.

Employees should be using AI to make their jobs easier, serve customers better, and fuel their innovation engines. In the public sector, this may require organizations to conduct a team skillset assessment. As with many industries, the public sector is unique when it comes to cultural and organizational considerations for AI. 

Data and Technology, Organizational Alignment, and People

Of these five dimensions, public sector entities need to consider three overarching components. In terms of data and technology, organizations need to ask themselves if they have the right systems, datasets, controls and governance in place to leverage AI in a safe and responsible manner. They also need to ensure their AI initiatives strategically align to their organizational goals.

Perhaps most importantly, they need to consider how they will deploy AI so that their workforce understands how to responsibly leverage the technology to accelerate the pace of their work to become more efficient and innovative, ultimately creating more value for the populations they serve. 

When AI is introduced without proper explanation and training, employees may go back to familiar manual processes they are more comfortable with. Some workers may perceive AI as a threat to their jobs, creating hesitancy or resistance to adopt. As such, successful AI deployment requires upskilling of employees to enable them to effectively and safely leverage
AI in their daily workflow.

Mitigating Risks

It’s critical for public sector entities to deploy AI in a way that prioritizes safety and responsibility across the organization. AI tools can easily sift through an organization’s entire data landscape, including personal information from the people served by the organization, accessing sensitive information without realizing it. This leads to a risk of inadvertently releasing AI-generated content containing classified or confidential information, posing significant reputational, financial, and legal liabilities to the public sector entity.

As such, public sector workers need to be able to recognize the potential risks associated with AI and understand how these tools can be used effectively and responsibly. As in other industries, public sector employees should avoid placing unwavering trust in AI generated content, and should instead hold a level of skepticism, conduct their own research, cite sources, recognize potential for bias, and verify the accuracy of content prior to public release. 

Although the U.S. government has some established AI guidelines,2 comprehensive state and federal regulatory frameworks are still in development. Organizations need to prepare for compliance with these future regulations. Noncompliance could potentially result in suspension of AI operations until they meet regulatory standards, resulting in an increase in time, effort, and resources. For this reason, public sector organizations need to align their AI initiatives with the latest governance and risk management practices to fortify themselves against future regulatory changes.

AI’s Promising Future in the Public Sector

AI can streamline the internal processes of public organizations and improve the quality of public services directly impacting residents and society. AI’s practicality extends beyond office automation of routine tasks like processing applications, issuing permits, and managing records. It can also be used to detect fraud anomalies in financial transactions and tax filings. Like other industries, many public sector entities are using AI-powered chatbots and virtual assistants to provide 24/7 support for resident questions. 

State and local law enforcement agencies are using AI for predictive policing to augment existing crime prevention efforts. AI is also being used in state and local elections. Some states use AI systems to enforce signature matching rules on absentee/mail-in ballots. In other cases, some states leverage the technology to improve transportation infrastructure, including AI-controlled traffic lights and AI-powered modeling, predictive analytics to understand how long road treatments will last, and bridge deterioration predictions.

Public sector entities can also use AI for environmental monitoring using satellite and sensor data, aiding in the prediction of a natural disaster. social service agencies can use AI to identify and provide services to vulnerable populations using predictive analytics.

These examples are just a small glimpse of the possibilities of AI applications in the public sector. Indeed, AI can transform public sector operations, making them more efficient, responsive, and data driven, with a direct impact on the communities they serve. But organizations need to embrace a thoughtful, deliberate approach as they start their AI journeys and consider using the AI readiness checklist to ensure sustainability for years to come.

JORDAN ANDERSON is a director with Baker Tilly’s digital solutions team.