The year 2026 will see major changes in the foundational infrastructure of the internet, driven more by political, legal, and physical challenges than by the latest advances in consumer technology. While AI and its demands on computing power will certainly make headlines, the real changes will be happening underneath, where issues of data handling, connectivity, and power grids will determine who gets ahead and who is left behind.

Four major trends will have significant impacts on businesses and the public in 2026. If I had to sum up the year in one sentence, it is this: 2026 is when the internet’s physical and legal borders start to matter again, and the companies that treat them as strategic assets will set the pace.

Sovereign cloud will evolve from a marketing term to a contract requirement

In 2026, expectations for sovereign cloud services will move past marketing slogans and become straightforward contract terms. Driven by updated regulatory frameworks in regions like the EU, contracts will specify who can operate a cloud service, where data will be managed, and which laws can apply to stored data. Triggered in part by recent high-profile outages, government agencies and critical sectors will write these demands into their procurements, forcing all providers, domestic and global, to prove compliance with explicit operational criteria.

Next year, we will see large buyers in places like Europe and parts of Asia include specific requirements for operator nationality, data handling practices, and legal boundaries into every cloud deal. Hosting locally won’t be enough. Clouds that are exposed to foreign jurisdictions won’t meet the standard of being sovereign. As a result of this shift, contracts for public sector, financial, and infrastructure workloads will not just ask where data sits, but also who runs and governs it. This trend will give regional and telco clouds a clearer opportunity to compete and set a new baseline for US-based providers aiming to retain customers under the updated compliance rules.

Neoclouds make connectivity, not chips, the new AI focus

Instead of focusing solely on acquiring more GPUs, fast-growing “neocloud” providers will compete with hyperscalers by controlling locations with strong connectivity and power, and clustering in carrier hotels and smaller metro hubs. Industry revenue estimates show neoclouds could grow from their $24 billion in revenue in 2025 to around $170 billion by 2030, reflecting their quickly expanding role in AI services.

As growth shifts to neoclouds, the fight is less about who has the most chips and more about where those chips sit and how the data gets between them. Providers who control premium fiber routes and reliable colocation sites in new markets will have more value to offer.

For enterprise buyers, 2026 will bring new products uniting compute and transport, so-called “AI corridors,” which will create advantages for carriers and colocation firms with dense cross-connects and predictable low-latency paths. This trend will shift growth to new markets like Milan, Warsaw, and Berlin, where infrastructure is less congested and grid access is more flexible, giving both local carriers and neoclouds an edge.

Power and grid access will control the pace of internet expansion

In 2026, electricity availability will become the biggest limiter for new internet infrastructure. Multiple studies point to steep power increases for European data centers, with some projections showing demand could triple by 2030. The practical impacts are already taking hold. In Ireland, new data center connections are paused until 2028 because data centers now use about 21 percent of the country’s electricity. In Belgium, grid requests from data centers surged ninefold in three years, prompting local governments to consider stricter allocation limits.

Most of the attention today is on chips and cooling systems, but the more immediate challenge is securing access to a stable power grid. From 2026 onward, companies choosing new data center sites will need power-mapping in addition to space or fiber connectivity. Building in major cities will become harder unless there is extra grid capacity.

The result is that secondary cities with spare electrical resources will start growing faster than established hubs. Expect a push toward direct power deals, greater use of liquid cooling to stretch any available energy, and data infrastructure changes that place new sites near renewable power supply.

Cloud and telecom regulation will overlap more closely

By 2026, the boundaries between telecom and cloud regulation are set to blur, particularly across Europe. Recent EU proposals hint that cloud providers and network services could eventually face similar compliance standards.

Following the October 2025 AWS global outage, discussions grew around treating hyperscalers as part of essential infrastructure, putting them under new resilience and reporting obligations. This means cloud providers will be held to standards historically reserved for telcos, such as incident disclosure and higher service guarantees, while telcos operating edge cloud platforms will need to adopt more advanced security and transparency practices.

By late 2026, I expect to see the first true ‘cloud carrier’ category made official, with companies requiring approval and oversight under both telecom and cloud regulations. This will put companies that already own networks and cloud services at a clear advantage since they can more easily meet overlapping requirements. For buyers, it will mean better consistency and clearer lines of accountability.

NORDVPN DISCOUNT - CircleID x NordVPN Get NordVPN [74% +3 extra months, from $2.99/month]