‘What’s our AI strategy?’ is the wrong question for agency leaders

Since the Trump administration unveiled its AI Action Plan in July, federal agencies have been processing the artificial intelligence innovation and infrastructure policy recommendations amid their struggles to navigate the influx of new opportunities to streamline ways of working. 

AI’s return on investment continues to be mixed or downright disappointing. According to a 2023 report from IBM, enterprise-wide AI initiatives achieved an ROI of only 5.9% despite 10% of budget allocation. 

The key to seeing real impact from AI adoption is to integrate AI into existing priorities, rather than treating it as a separate strategy, disconnected from the organization’s real objectives. For agencies, asking “what is our AI strategy?” misses the point. 

As the AI revolution continues and mandates emerge, federal leaders should be asking “how does AI enable our agency objectives?” This will ensure that AI adoption accelerates and enhances existing efforts designed to meet organizational goals. 

Pushing for adoption that isn’t rooted in overall objectives will make it harder for employees to embrace, because the effort can feel disconnected from the agency mission and culture. Here’s what to do instead.

Start with people, not platforms

More than any past digital transformation, AI’s success depends on human judgment and behavioral change. Past digital transformation efforts automated or changed workflows in ways that once adopted, didn’t require people to make constant decisions. Once implemented, the software was a finalized structure, and users had largely no choice in how it was used. 

AI in its current form is instead molded by our prompts, feedback, and usage patterns. There is no “one way” in how we use the majority of AI tools. Generative AI, which is rising in popularity, can be used by two employees with identical objectives and access, but with completely different outcomes. 

Leaders navigating this complex adoption can inspire the right uses and behaviors by clearly articulating how AI integration will improve the agency’s mission. Otherwise, the benefits of AI will be limited to a tool that can improve some processes but will remain unlikely to transform how work gets done.

This is challenging, especially within the widespread media focus on job losses from AI, resulting in fear and anxiety. Employees need some clarity on how their day-to-day work and long-term career trajectories will be impacted. AI should reduce administrative tasks and increase productivity, not create stress and confusion. 

Agencies need to communicate the “why” as clearly and precisely as possible, while showing empathy and a willingness to learn and adapt to help deter any fear that can stifle participation. Creating environments that encourage testing to make employees feel safe to experiment and learn is also paramount, instilling confidence in workers and accelerating adoption. And finally, agencies should reward behaviors that are in line with this expectation of experimentation and learning.

Adopt a multi-channel approach

AI integration works best when top-down strategic alignment works together with bottom-up experimentation. Senior leaders should map out their AI initiatives to directly tie in with the organizational business strategy, providing clear objectives and allocating resources in alignment with their agency’s mission. If employees can’t see how AI correlates to their core goals, it will be hard to understand why it deserves an investment. 

At the same time, it’s important to acknowledge the reality of where AI adoption stands today. The technology is still in its early stages, and the agencies that make way for experimentation and view AI as a learning process — not a one-off deployment — will see the most success.

Success will require listening to input from frontline teams who have insights into precisely where barriers and inefficiencies can be addressed through AI. Leaders should encourage employees to take initiative to identify solutions, creating a sense of ownership for the change. This in turn helps teams see AI as a way to improve work, not simply a threat to jobs. 

Combining strategic direction from leaders who have a more holistic view of the organization — along with insight from employees who understand what it takes to get there — will maximize the potential of AI adoption. 

Practice adaptive leadership

Success with AI relies on testing, learning and adjusting in real time — something that traditional governance structures built for predictability typically resist. 

To encourage the kind of experimentation and initiative that is required, leaders will need to welcome ambiguity, speed, and prioritizing agility over control. This translates to removing layers of approval, creating dynamic and experimental teams, engaging people across the agency in finding solutions, and being transparent in communications. 

The AI Action Plan is an opportunity for agencies to accelerate their progress toward ongoing objectives. Taking advantage of this opportunity requires an outcome-focused approach rather than an activity-focused one. 

To capture the potential of AI, leaders will have to root all initiatives in the overall strategy, engage employees as partners in change, and build a culture of adaptability. The true transformation happens when employees understand, trust, and embrace AI implementation plans. 

Human behavior is at the forefront of successful AI adoption. If your workforce isn’t equipped or motivated to use AI effectively, you are unlikely to see real value generation. 

Gaurav Gupta is the managing director and head of R&D at Kotter.