At the first ever CITO East Virtual Africa Summit, Dr Leah Miano, head of technology at Actionaid International, observed that AI discussions are slowly shifting to value realisation. She urged organisations to invest in governance and staff readiness if they are to gain economic value from AI.
During the CITO East Virtual Africa Summit, held on November 5, Dr Leah Miano, head of technology at Actionaid International, observed a shift in AI conversations in the last three years that is forcing organisations to think deeper on the economic value of AI adoption.
“It’s not about whether we can use AI, but under what conditions we can use it safely, responsibly, and confidently to be able to achieve business value,” she said.
In her address, she described to attendees the conditions she believes are necessary for successful AI adoption in the modern business environment.
The first condition she believes is a necessity is governance. She argued that governance must not be viewed as a barrier to innovation, but rather as an enabler. To do this, she encouraged attendees to always ask themselves three questions when trying to fix governance in their organisations.
“Who is accountable for the AI outcomes in your organisation? Do we know where AI is being used, including the unofficial AI, and can we demonstrate to our boards that the AI initiatives are governed, secure, and delivering value?” she asked.
Fixing governance, she said, enables organisations to scale to greater lengths. However, she encouraged the nearly 100 tech leaders gathered in the room to adopt compliant AI practices into everyday operations from the get-go as opposed to later.
A second condition that must be met for successful AI adoption, according to Dr Leah, is performance measurement. She encouraged tech leaders to establish a return on investment baseline and articulate the total cost of implementation.
“The conversation needs to move from how many AI tools we’ve deployed to what has fundamentally changed in the way organisations operate,” she said.
She further explained that AI investments must be measured against an operational baseline, cost savings, improvements, and error reduction, especially because they are not cheap.
Staff readiness and change management
Besides measurement of performance and governance, organisations must also invest in staff readiness and change management. For Dr Leah, successful AI adoption must be preceded by preparation, which includes practical AI literacy, learning opportunities, and clear, acceptable use guidelines. Preparation then necessitates a change management process so that every new change is aligned to business processes.
“We must partner very closely with business functions so that it doesn’t remain like a technology thing, to be able to redesign their processes and proactively manage their impact into evolving roles and responsibilities,” she said.
She also addressed the role of data in successful AI adoption. While she recognised that data has become an important part of the AI conversation, she warned that AI can only perform based on the data given and can’t compensate for poor data foundations. As a result, she emphasised that before organisations can scale AI models, they must establish clear data ownership, which then helps determine who owns the AI outcomes in an organisation. She further advised leaders to maximise enterprise platform capabilities before procuring any redundant point solutions.
With increased adoption of AI agents, Dr Leah noted that workflows are going to become increasingly autonomous.
“The next phase of enterprise AI will move from being isolated, task-based content generation towards orchestrated end-to-end workflows that make and get business value,” she said.
With that transition, human intelligence will set the direction, and the agents will do the work. As a result, Dr Leah urged IT leaders to make sure that transition is governed, secured, and focused on organisational outcomes.
“Start with a problem, not with a solution, so that it makes organisational sense. AI agents offer significant potential to complex activities across improved and approved enterprise systems, and the ultimate goal is not about chasing autonomy just for its own sake, but to achieve better operational outcomes without compromising accountability,” she said.





