artificial intelligence


How to get beyond AI myths

It is the year 2020, and Michael is a senior IT manager who is about to lead a deployment of the latest version of Microsoft Windows to over 10,000 government computers. He and his team will be trained by Microsoft to support the oncoming surge of help desk tickets that usually come with these upgrades. The training would be provided by Microsoft’s latest artificial intelligence training solution, which, in testing, seemed quite impressive at answering even the most technically complex questions quickly and accurately.

During the actual training, however, Michael felt restricted because he would have liked to interrupt the training and discuss with his colleagues exactly how certain aspects of the new operating system will behave in their environment. He found himself making specific notes for his team during training because he felt there was a risk that the rest of the team was not picking up on the nuances of how the operating system will behave. While the AI training solution was incredibly efficient and flexible allowing the team to take the training almost anytime and anywhere, it did not allow the kind of collaboration and collective learning that would ultimately result in the IT team learning how to apply the program's specific aspects to their users.

How much is too much AI? What are the considerations for today’s IT manager who stands to benefit from automating highly repetitive but “intelligent” tasks or activities? If and when should a human intervene? And what is the degree and type of communication necessary to realize savings and effectiveness? These are questions for today's AI experts both in government and private industry.

McKinsey, a large strategy consulting firm, purports that there is large potential for AI to contribute to economic activity, but that adoption of AI could widen gaps among countries, companies and workers. The analysts predict that:

  • By 2030, 70 percent of people might have adopted at least one type of AI.
  • China will be the global leader in AI and is investing heavily now.
  • Jobs could shift away from repetitive tasks to those that are more socially and cognitively driven and require more soft/digital skills.

When we see all the "AI hype" in the media and learn about the different AI applications that organizations are adopting, it is easy to jump to conclusions about how AI can increase organizations’ efficiency and create a pathway to success. Four of the most common conceptions sound reasonable at face value; yet upon considering the different and complex systems in which we operate, we can declare all four of them as myths.

Myth # 1: Integration of AI into an organization’s business process is a guaranteed path to success.

The only guarantee to success in implementing any change is a unified purpose and well-developed strategy. A clear purpose for the use of AI must be defined and agreed upon prior to implementation if an organization is determined to make a sustainable change. Leadership should engage in a robust strategic planning phase with a targeted consulting intervention help the organization define its vision for AI implementation and develop a strategic plan that has the needed buy-in from all strategic players. Employing a third party to conduct the intervention is helpful to reduce bias in the process.

Myth #2: AI will help organizations be more efficient and effective.

Managing the implementation of a new technology  as complicated as AI is not the real challenge. We believe that managing the people involved in the process is where the challenge lies. A common belief is that AI will revolutionize the way people do their work; hence, managing these processes and practices will ensure a successful change.

A human-centered approach, however, is critical in moving toward AI. The technology will not replace people’s jobs, but it could transform them. A restructuring or redesign of the workforce will be needed to meet new working requirements, which should be governed by well-designed and standardized processes across multiple functions. The perfect toolset to use here is a human-centered business process improvement capability that will align newly structured processes to new roles and responsibilities, greatly reducing risks around achieving strategic priorities enabled by AI.

Myth # 3: AI will help people communicate better.

AI might help enhance the communication platforms available for people, but it will not improve communication on its own.  Effective communication within any organization is always driven by leadership's guidance and commitment. In fact, leaders should be very precise and intentional about setting guidelines for when AI communicates with people, especially customers.

When it comes to communicating within the organization, leadership development can help prepare executives for the changes their organization will be facing with the implementation of AI. Leaders must be deliberate in their direction emphasizing communication to ensure a successful and sustainable change. Process leaders must empower employees to identify the hard decisions that people need to make because AI will fail to do so. Employees should feel encouraged to step in and lead communication with outside stakeholders when appropriate to ensure effective decisions are being made.

Design thinking is an excellent approach that can be used in remodeling the communication plan so that customers are at the center and people’s experiences are closely aligned with the original goals of the AI implementation.

Myth # 4: AI will handle ethics and safety.

Complex decision-making processes go beyond choices of when people should intervene to communicate directly with others. Ethical standards, safety considerations and emotional choices have always been what set people apart from machines. People are “equipped” with sophisticated value systems that help them navigate the complex decisions. Organization development and behavior concepts and methodologies can help teams align their value systems with the purpose of the change their organization is about to face through an AI implementation. Team cohesion and shared purpose ultimately guide people to make effective, complex decisions that build resilience against new challenges and unexpected changes. In the end, a high performing team that implements a human-centered change management process is what will ensure a sustainable AI solution.

Many gaps can be identified between what we imagine our future to be with AI and our current state of determining how to derive real value from the technology. These gaps are created by unclear boundaries of when human decisions and intervention should override AI use and assumptions that simple implementation of AI will fix all our problems. To achieve the full potential of the future we desire, organizations should continue to focus on the people who will enable technology to serve us. Leading with a human-centered approach, executives should understand the importance of leveraging strategic planning, business process improvement, high performance teams, human-centered change management and organizational development approaches to successfully maneuver a change as complex as AI implementation.

About the Authors

Emad Elias is a principal with Evans Incorporated.

Rasha Fakhreddine is currently working as a consultant for Evans Incorporated.


  • business meeting (Monkey Business Images/

    Civic tech volunteers help states with legacy systems

    As COVID-19 exposed vulnerabilities in state and local government IT systems, the newly formed U.S. Digital Response stepped in to help. Its successes offer insight into existing barriers and the future of the civic tech movement.

  • data analytics (

    More visible data helps drive DOD decision-making

    CDOs in the Defense Department are opening up their data to take advantage of artificial intelligence and machine learning tools that help surface insights and improve decision-making.

Stay Connected