The fast-tracking of cloud computing continues. According to Gartner, the worldwide public cloud services market is expected to grow by 20.7% to total $591.8 billion in 2023, up from $490.3 billion in 2022.
As the CTO of BT, my team and I are embedded in the ongoing digital transformation  of our enterprise customers and the shift of their apps and digital workloads to the cloud. There, they benefit from the scalability and cost efficiency of hyperscaler platforms, pay-as-you-go billing with no up-front capital costs, robust disaster recovery and, turn-key security. This, combined with improved accessibility, collaboration and productivity can help boost productivity and accelerate innovation.Â
However, after early successes, many customers now face digital growing pains as more and more of their apps and workloads move to a multi-cloud environment and the complexity of managing it all grows.
Let’s explore four key trends enterprises should be aware of when it comes to the transition to multi-cloud computing.
Adoption of a zero-trust approach to security
With the rise of cloud computing comes increased concerns about security—particularly with today’s distributed workforce and growing number of connected IoT devices. It’s no longer easy to see which assets are corporate-owned and who ‘your’ employees really are. The traditional perimeter security model of ‘trusted inside’ versus ‘untrusted outside’ no longer exists, particularly in a multi-cloud environment. A zero-trust approach to security is now needed. This assumes that all network traffic is untrusted and requires authentication and authorization at every step.
Companies need to prioritize cloud security and ensure that data is protected across all cloud environments. Hybrid and multi-cloud architectures may also require changes to an organization’s IT infrastructure, including network architecture and security protocol.
Taking a data-centric focus to security means building a dynamic, policy-driven model that changes in line with the attack surface.
Even in the most thorough network architectures, trust should never be assumed. Users and devices need to always be verified and continuously monitored. The identity of devices, users, data, and assets are the new perimeters, so control is enforced as close to these as possible.
The rise of managed distributed clouds
In BT’s recent global research study, 91% of business and IT leaders said improving data security and data sovereignty was a likely technical reason for upgrading their IT and network infrastructure.Data sovereignty — the extent to which data is subject to the laws of a country, no matter where it is stored — can be quite complex because some countries mandate that data and cloud services remain off-limits to foreign actors, be they companies or governments. In fact, three out of every four countries have some kind of regulation around data sovereignty or localization of data according to a McKinsey & Company study.
To meet their growing demands, more organizations are turning to a distributed cloud — an architecture where multiple clouds are used to meet compliance needs, performance requirements, or support edge computing while being centrally managed by a trusted partner working directly with the public cloud provider. At its core, a distributed cloud service is one that runs in multiple locations It needs to provide interoperability across multiple cloud infrastructures and smooth paths for data exchange. And it needs to be efficient and not lead to excessive egress and bandwidth charges or resource-heavy API integration.
When data sovereignty regulations state that data can’t be moved to a public cloud provider, a managed distributed cloud can effectively be moved to the data. This can help companies comply with governance and regulatory mandates and means data can be processed efficiently with the minimum amount of latency. It’s a win-win.
Gravitation towards energy efficient hybrid and multi-cloud architectures
Data traffic is predicted to grow eightfold by 2030 (BT & Accenture). This could leave a significant carbon footprint depending on the source of energy fueling that data. At the same time, businesses are also under pressure to be more agile and go to the cloud. The problem is that everyone’s infrastructure is different, and no two cloud journeys are the same.
Cloud providers are competing to attract customers—sometimes having to court different departments within a customer organization—to land their application, data set or service on their cloud. This may be driven by technical compelling events, but there are also disruptors such as mergers, acquisitions, and divestments. Overnight, an organization could become multi-cloud once a deal lands.
Furthermore, many application business cases are multi-year investments heavily weighted on upfront development of intellectual capital. Hyperscalers are excellent for reducing operating expenses and consumption models, which ultimately drives innovation. When the cloud application business case moves to a maintain-and-run state where data volumes have increased, those applications become targets for optimization. Hybrid cloud or an alternative cloud provider has an opportunity to compete for that business on a total cost of ownership reduction basis.
If Moore’s law is applied here, there are sustainability concerns that get addressed through modernization while providing access to additional budget. Physical footprints are reduced along with consumption of power while there’s a recognized increase in usable data processing capacity. Furthermore, leveraging services that are contained in energy-optimized accommodation allows organizations to address their sustainable supply chain commitments.
Increasingly organizations are looking for energy-efficient multi-cloud services to align with their sustainability goals. They are looking for the optimum balance in their use of public and private clouds of  performance, cost, and security. Commercially, it mitigates risk by diversifying cloud investments and forcing greater competition for business between the cloud providers.
This trend is important because it allows organizations to have greater flexibility and control over their cloud infrastructure.
Integration of AI into cloud computing
A survey by IDG found that 87% of IT decision-makers believe that AI and machine learning will be important use cases for cloud computing in the future. There is a synergistic relationship between AI and cloud computing, and they can work together to enhance each other’s capabilities. AI can be used to automate cloud computing tasks like resource allocation, load balancing, and scaling, helping optimize system performance and improve user experience. Meanwhile, cloud computing provides the infrastructure and computing power AI models require so they can train and deploy at scale.
AI can also detect and prevent security breaches in real-time because it’s constantly analyzing patterns and user behavior. And because AI can predict behaviors and usage patterns, it can also offer organizations cost-saving benefits, knowing how to allocate resources and not over- or under-commit resources.
From taking a zero-trust approach to security to understanding how to function efficiently amid increased regulation to moving toward more sustainable, energy-efficient cloud computing, the industry is evolving at an unprecedented pace. Look for these key trends to continue their importance through the remainder of this year with advancements on the horizon.
