The technology industry is in the midst of a skills shortageone that shows no signs of slowing. The U.S. Bureau of Labor Statistics projects that tech jobs will grow at twice the rate of Americas overall workforce, creating hiring shortfalls as organizations struggle to fill critical positions in IT, cybersecurity, and other vital areas. The emergence of AI has only exacerbated the issue, as organizations in nearly every industry are seeking employees who can help them better understand the technology and get the most out of their solutions. Even as AI becomes a part of everyday life, most organizations are still determining how best to utilize itand how to limit the risks it may pose.
Interestingly, these challenges mirror another (relatively) recent innovation: the cloud. Before cloud computing became commonplace, businesses with any sort of digital footprint needed to buy rack space or manage their own on-premises servers. That was reasonable for businesses with a high degree of technical expertise, but building and maintaining a climate-controlled server room wasnt realistic for most companies. The advent of cloud computing democratized access to advanced computing capabilitiesand AI is already having a similar impact. As businesses wrestle with managing and securing their AI deployments, they can look to the cloud for lessons and guidance on how similar challenges were tackled in the recent past.
The Evolution of Cloud Adoption and Security
Consider the cost of on-premises computing. Server rooms are expensive, as are the servers themselves. They also require a substantial degree of technical expertise to maintain, and employees with experience in that area understandably command high compensation. The emergence of platforms like Amazon Web Services, Microsoft Azure, and Google Cloud changed all that, lowering the barrier to entry for advanced computing capabilities: businesses could eliminate the high initial investment associated with purchasing servers and building server rooms in exchange for a modest (in most cases) operating expense. Perhaps most importantly, it allowed businesses to work with reliable partners to get the most out of their cloud services, rather than relying on difficult-to-come-by in-house expertise.
That said, securing the cloud still presents its own challenges. During the early days of the cloud, businesses often made the mistake of assuming that providers would handle any cybersecurity needsa misconception that left them dangerously exposed. Today, the most common rule of thumb is that the provider is responsible for the security of the cloud itself, while the customer is responsible for the security of the data inside it. Essentially, the provider ensures attackers cannot exploit their systems to get to your data, but if poor password management, device security, or other data hygiene practices allow attackers to compromise your accountsthats on you. This delineation has helped businesses better understand where their potential risk factors lie when it comes to cloud security and mitigate them appropriately.
Applying Lessons from the Cloud to AI
Its not hard to see the parallels between the cloud and AI. Like the cloud, AI has democratized access to resources that were previously difficult to come by for many organizations. The widespread availability of generative AI models like ChatGPT means organizations no longer need to hire costly AI engineers to create, manage, and fine-tune their own models. Instead, they can put an application layer on top of an existing model and deliver a compelling service to their customers at a relatively low cost of ownership. While this still requires a level of technical expertise, the barrier to entry is much lowerand organizations can move forward faster with smaller, more flexible engineering teams.
The risks posed by this model mirror those posed by the cloud. When you upload data to the cloud, it is no longer under your direct control. The same is true of third-party AI modelswhen customers (or employees) input data into an AI-powered application, its important to know where the data is going, how it is being stored and protected, and how it is being used. With AI still in its relative infancy, the answers to those questions arent always clearwhich means businesses providing AI functionality need strong AI governance practices in place to establish trust with their customers. For some businesses, that might mean offering customers the ability to opt out of AI features. For others, it might mean putting clear safeguards in place to prevent AI tools from accessing sensitive or confidential information.
By demonstrating the ways in which they are limiting AI risk, businesses free their customers to evaluate the benefits of AI. Over time, most businesses migrated to the cloud because the efficiency gains substantially outweighed the perceived risksand a similar pattern is already emerging when it comes to AI. In fact, its happening even more quickly this time. Since AI has use cases across nearly every business unit, the potential ROI is much easier to illustrate. While its true that every customer will have a different risk appetite, the trend is clear: eventually, nearly every business will decide that the rewards significantly outweigh the risks. By establishing strong governance practices and lowering the amount of risk associated with AI, businesses can help their customers reach that point more quickly.
Freeing Customers to Embrace AI with Confidence
While businesses and their customers are understandably concerned about AI risk, the history of the cloud provides a helpful road map for navigating those risks successfully. The risks associated with AI are not dramatically different from those associated with other technologyand businesses can mitigate them in much the same way. By establishing strong governance practices and implementing clear transparent policies regarding AI and its use, businesses can enable their employees and customers to embrace the potential of AI with confidence.
