Navigating cloud concentration and AI lock-in

It’s not that difficult to understand. When we move to any new technology, we increase risk. Risk of lock-in, failure, and, most commonly, the risk of building solutions that don’t return enough value to the business.

Gartner surveys have underscored the emerging risk of cloud concentration. They emphasize the potential wide-scale impact of business continuity failures associated with overdependence on a single cloud provider.

This is one of the most frequently asked questions from my clients moving to the cloud, and for good reason. They already saw what occurs when you put too many eggs in a single enterprise technology basket, meaning lock-in and overdependence on that specific company. They don’t want to repeat this mistake as they move to cloud-based platforms.

This risk is further compounded by the growing concern over the emerging AI tools market and the accompanying risk of AI lock-in. CIOs may find switching between AI technologies from different cloud vendors challenging and costly. As I often hear, anything is possible with enough time and money, but we don’t have much of either these days. The worry is that we’ll move in the wrong direction. In the past, this meant spending more money to correct course, but given the strategic value of artificial intelligence, taking the wrong path could mean shutting the business or, more likely, selling it cheaply.

Weighing your options

One of the main concerns IT leaders grapple with is the control of AI platforms by the big hyperscalers. The competition among major cloud providers to establish dominance in the tools market for AI and generative AI has spurred an arms race. As I’ve pointed out a few times, the big cloud conferences have become generative AI conferences, with cloud computing a secondary topic.

This potentially limits CIOs’ ability to access and utilize the best-in-class AI technologies from different vendors. What happens when the superior AI technologies don’t happen to be native to the primary cloud provider with whom the CEO made an exclusive deal five years ago? Whoops.

Gartner’s reports have ranked cloud concentration as a significant emerging risk, and I agree. We need to consider the implications of third-party viability, evolving sociopolitical expectations, and the availability of mass generative AI. We’re moving too fast in the same direction, which could exaggerate the impact of the wrong decisions, including focusing only on a single cloud provider and their native AI solutions.

This can also bring down the business. The stakes are higher than just another down year due to dumb leadership mistakes. The climate is unlike any I’ve seen in my long career. Board members and investors are reaching out to me and other subject matter experts to manage these risks.

What to do?

Furthermore, the domination of large public cloud providers has created a gravitational pull that makes migration out of a single cloud provider costly and labor-intensive. As a rule of thumb, it’s going to cost you twice as much as your initial investment to either repatriate back to traditional systems or move to another cloud provider. This is thanks to the cost of talent and the fact that you’ll be working on two platforms for an extended period.

Although you can choose to reduce the use of a specific cloud provider, it is sometimes nearly impossible to move some applications to other platforms. This is due to the coupling of those applications to the cloud platform and the economic inability to get them off those platforms.

To guard against the risks associated with cloud concentration and AI lock-in, IT leaders are exploring strategies to reduce dependency on a single cloud provider. This can include leveraging single-tenant cloud solutions, colocation companies, and hybrid cloud strategies to diversify their cloud deployment and infrastructure.

As IT leaders navigate the complex landscape of cloud concentration risks and AI lock-in, it is evident that an agile approach to cloud strategy and AI adoption is mandatory. Organizations can mitigate risks by understanding the nuanced considerations of vendor selection, fostering a multicloud approach, and embracing innovative technologies. At the end of the day, keep your eyes open for the fully optimized solution, and do not focus on just a single cloud provider’s services, including AI.

My advice is much the same as adopting any group of technologies. A strategy is needed first in terms of what the business should be and its most desirable state. Then form a strategy around how things need to change, including the whys, hows, and whens. There won’t be a desired end state, but rather many end states that need to be reached as the business evolves and thrives.

Keep in mind that a strategy is not the same as a plan. You’ll still need an overarching plan defining exactly what needs to be done, including resources, timelines, processes, and the ability to be agile through those processes. Many things will go wrong, and you need to be willing to take a step back and correct them. The businesses that can do this will survive past the next 10 years. Try to be one of them.