We are all experiencing the rapid proliferation of Artificial Intelligence in the workplace at an accelerated pace and how it’s elevated to the top of executive boardroom agendas, globally.
Customer service-based organizations are deploying AI assistants and virtual agents to handle routine inquiries, resolve common issues and provide 24/7 support. Software development and IT operations-based organizations are using AI-powered coding assistants to help developers write, test and review code faster. Supply chain operations are using AI to help forecast demand and identify operational inefficiencies.
While there are use cases, by and large, where some organizations have successfully adopted AI into their work forces, they are still few. The reality is that many organizations have experimented with AI, but very few have successfully scaled it across the enterprise.
But why?
Executives are no longer challenged with “should we use AI?” but rather “how do we use AI safely, responsibly and at scale?”
In contrast, some see the application of AI guardrails and governance an impedance to the velocity of AI adoption and innovation.
This article will discuss the adoption and governance lessons we have learned from past technology introduction experiences and share real world case studies on how we can drive AI adoption velocity at scale.
For those of us who remember, there is a familiar technological experience we can learn from where we can draw parallels from: The Internet Use Policy.
Yes. Almost every organization, who could afford broadband internet, had one. With the introduction of the internet in the late 1990’s, large enterprises faced concerns remarkably similar to those that risk-averse businesses face today with AI. Without the appropriate guardrails, employees could access inappropriate content, sensitive corporate information could leave the organization, malware and viruses could enter the network, intellectual property could be exposed.
Businesses’ response was not to ban the internet but to put the necessary policies and technological guardrails in place. The internet use policy was born, security standards were drafted, security technologies were implemented, like firewalls, content filtering and data protection controls.
The conversation quickly converted from “the internet is too risky, we cannot allow employees unrestricted access”, to “with our internet use policy and technology controls in place for monitoring, we can allow broad access because we have put accountabilities in place”.
The same pattern is emerging with AI. Many organizations are concerned about AI exposing sensitive data, generating accurate information, unknown impacts to regulatory risk particularly when autonomous AI systems make decisions that may be difficult to explain.
Another technological experience we can draw a very similar parallels, again, for those of us old enough to remember (wink): Consumerization of IT.
Consumerization of IT was one of the defining technology management debates in the 2000’s and through to the 2010s. I remember reading IT magazine articles full of perspectives around bring-your-own-device (BYOD) policies. Before the 2000s, companies typically select, approve and deploy technologies for their employees but with the dawn of competing handheld technologies like Nokia, Motorola, Blackberry and Apple; consumer internet services and laptops, employees began bringing their technologies already being used for personal use into the corporate workplace.
But what drove this behavior?
Employees often had better technology at home than what companies were offering. Plain and simple. I remember making a case for my Blackberry Torch (Blackberry’s first attempt at the touch screen) at my former company, mainly to show off my new personal tech at coffee break but I was denied and ended up with a Qualcomm mobile phone.
Similar to the introduction to the introduction to the internet, many organizations tried to block or restrict employee driven technology adoption. Instead, while businesses foresaw that the trend could not be stopped, enabled mobile device management (MDM), implemented BYOD policies, and introduced identity and access controls, data loss prevention and other like security-based technologies.
Just as employees began bringing personal devices to the workplace before formal policies existed, in the age of AI, many employees are bringing their ChatGPT, Claude, Gemini, Microsoft Co-pilot to work and the likes to the workplace , bring your own AI (BYOAI), regardless of whether formal AI governance frameworks exist or are mature.
What we learned from the Consumerization of IT and Internet Use Policy eras, that initially, organizations viewed governance as an obstacle to advancing the adoption of technology, but history suggests the opposite.
Now let’s look at AI governance. There are many perspectives from risk and security professionals on this topic and although the topic to broad stretching, the one prevailing attribute to protect is privacy.
A former Canadian Privacy Commissioner, Dr Ann Cavoukian, widely known for her Privacy by Design concept. Introduced as acting Privacy Commissioner in the last 1990s and widely applied throughout her tenure into the 2010s, the framework dictated that privacy must be proactively embedded into the design of IT systems, business practices and network infrastructure (. In my former role as an Account representative for the Health and Government sector, our executive leadership mandated that we learn and apply the seven foundational Privacy by Design concepts (footnote www.ola.org) to assist our privacy federally regulated customers with their compliance requirments.
- Be Proactive not reactive
- Privacy as the default setting
- Privacy embedded into design
- Positive-sum, not zero-sum (operating without unnecessary privacy trade-offs)
- End to end security
- Visibility and transparency
- Respect for user privacy
What Privacy by Design taught us is that governance should not be viewed as an impediment to innovation but to make innovation safe, trusted and scalable. In a way, Privacy by Design is not just a privacy framework but should be revisited by organizations as a blueprint for responsible AI adoption.
With governance, enterprise-wide adoption becomes possible, and businesses are more free to innovate because guardrails and accountabilities were put in place.
While the past has taught us that governance is not a hinderance but an enabler, let’s tackle how businesses can learn how to scale AI.
I am fortunate to be in the position where I work with enterprise clients and learn how they drive AI adoption in their organizations. Before you can unlock AI adoption velocity, you need to first find your AI champions. One conversation stands out with a client of mine where this regulatory based firm, applied a gamified approach. The firm invited teams of participants to develop AI solutions through a hackathon themed innovation workshop. The hackathon was built around a formal learning session as well as an adoption journey. Participants first learned about AI agent fundamentals, then enablement sessions which then moved to the competition where teams were tasked to build working AI agents, which were later judged and awarded. While the solutions that were built are not the prime argument for this example, the real value was the adoption benefits that it generated. In a feedback loop from the event, participants shared a reduced fear of AI development and the structure of learning before developing helped develop internal AI skills and eventually inspired some with their career paths in AI at the firm. The Hackathon was the adoption mechanism.
“The way to conquer fear is through understanding” Ralph Waldo Emerson
For unlocking the speed of AI adoption, I’m going to source my key learning from “Artificial Intelligence for Leadership by Danielle Stouffer. The author illustrated this point through a case study based on a concept: Decision Velocity Metrics. In the use case, this firm could not deploy AI solutions any faster than 4-6 months. The AI lead examined their entire enterprise process from start to finish and pointed out numerous use case iterations, reviews, demonstrations, business level approvals, executive level approvals and hand-offs. The challenge was that the team could only advance through each approval when specific criteria were met. The decision velocity metrics tracked where the delays were occurring and why. The result was that it was not about working faster or adding resources but removing the barriers. This transformation resulted in the next AI implementation in just 97 days.
Removing their barriers was not only a path to unlocking the velocity and scaling AI but also became their biggest competitive advantage.
If history teaches us anything, it is that innovation and governance are not opposing forces, they are partners in progress.
The Internet did not become a transformational business platform because organizations ignored risk. It succeeded because businesses developed policies, security controls, accountability frameworks, and standards that enabled trusted access at scale. Consumerization of IT did not reshape the workplace because organizations surrendered control. It succeeded because leaders adapted, modernized governance, and found ways to embrace employee-driven innovation safely.
Today, AI represents the next chapter in that story.
Today’s concerns are real. Data privacy, model accuracy, bias, explainability, regulatory compliance, and autonomous decision-making are challenges that deserve thoughtful attention. Yet these concerns should not be interpreted as reasons to delay adoption. Rather, they are signals that organizations must build the same foundations that allowed previous technological revolutions to flourish.
The lessons from Internet Use Policies, Consumerization of IT, and Privacy by Design are clear: governance is not the destination, it is the bridge. It creates the trust required for innovation to move from disparate experiments to enterprise-wide transformation.
Organizations that view governance as a barrier will likely continue to struggle with fragmented AI initiatives, isolated pilots, and limited business impact. Those that view governance as an enabler will create the confidence needed for employees to experiment, leaders to invest, and businesses to scale.
Perhaps the most important lesson is that technology adoption has never been primarily a technology problem. It is a people problem. Success has always depended on education, trust, accountability, overcoming the fear of the unknown and the willingness to adapt. Whether through hackathons that reduce fear through hands-on learning, AI champions who inspire their peers, or streamlined decision-making processes that remove organizational friction, the path to AI adoption is ultimately paved by people.
The good news is that we are not navigating uncharted territory. We have decades of experience introducing disruptive technologies into our organizations. We know how to create guardrails. We know how to manage risk. We know how to build trust. And we know how to scale innovation.
The question is no longer whether organizations can adopt AI safely and responsibly.
History has already shown us that they can.
About Fergonn
Fergonn Fernandez is a trusted advisor and strategic partner to Canada’s leading Banking, Financial Services, and Insurance (BFSI) organizations.
Fergonn brings a deep understanding of the Canadian financial services landscape, regulatory environment, and emerging trends including AI, automation, and cloud. He focuses on helping clients accelerate growth, modernize operations, and deliver exceptional value to their customers.
