Another year is ending, with another about to start. Good timing then for SDxCentral to stake its bets on what will define the technology news cycle in the year ahead. Besides much bandied AI agents, of course.

Spanning networking, telecoms, supercomputing, cybersecurity, AI, and more, the following predictions are what we believe will define network news in 2026.

Expect editorial opinions on 6G madness, the consolidation churn, sovereignty hype, and which embattled tech giant will likely have its day in the year to come.

And perhaps a little nod to agentic AI, with apologies.

AI will finally break the internet

There’s a lot of AI out there, but we’ve yet to see a purely AI-related attack to make the headlines.

What’s happening now is pretty obvious: AI is helping hackers reduce time to attack, and helping attackers whip up malicious code. Some on the dark web are touting exploit-kit-as-a-service (EKaaS) platforms that use AI to autonomously profile target systems and identify vulnerabilities.

What we haven’t really seen – yet – is a large language model (LLM) or agentic-led attack. But the stirrings are there.

Recently Google claimed to have found and stopped the first observed live malware, Promptsteal, which employs an LLM to generate on-demand Windows commands for data exfiltration and system reconnaissance.

Another worrying threat, also neutralized by Google, is Promptflux, which can replicate itself by copying onto removable drives and mapped network shares.

So network security has to watch out in 2026. The same goes for network resilience teams: both Cloudflare and Nokia are already warning of ‘terabit tsunami’ attacks which AI will only help galvanize.

You remember those outages this year? Expect a lot more in future.

- Giacomo Lee, News Editor

Intel to go to the moon

From languishing in late 2024, Intel will bounce back. A bold call. But with its Panther Lake processor – the first built on its advanced 18A manufacturing process – now shipping, it provides an early and much-needed positive step in the right direction.

Could it close the gap on Taiwan Manufacturing Semiconductor Co. (TSMC)? Not likely, but that won’t stop its foundry business from gaining some real traction from customers wanting an alternative to the already stacked Taiwanese firm, but also one that manufactures in the U.S.

Meanwhile, after several quarters of malaise, Intel finally returned to profitability during its fiscal-third quarter of 2025 after losing nearly $19 billion in 2024. All that aggressive cost-cutting last year helped to massively reduce its workforce, while CEO Lip-Bu Tan has sought to bring fresh discipline and focus from the boardroom down, rather than trying to compete with juggernauts like Nvidia – a move that culminated in the latter’s $5 billion into Intel.

Its stock prices are already surging compared to its 2024 figures, too – setting investors up to potentially provide some more much-needed confidence in the fledgling firm. If Intel executes on 18A yields, converts its foundry pipeline into actionable revenue, and captures apparent AI PC growth, 2026 could see significant multiple expansion as skeptics turn believers.

- Ben Wodecki, Senior Editor

We can’t quit cyberattacks

Despite best (?) intentions, cybersecurity attacks on network infrastructure will continue to proliferate in 2026.

Now, this might not be the most astounding prediction, but it should be.

We have been bombarded for years (decades?) about industry changing advances in cybersecurity technology and deployment models, only to time-and-time-again see cyberattacks cripple critical infrastructure.

Whether that’s because those new cybersecurity technologies are not all they are cracked up to be or are too hard to implement, or that the cyberattackers are always able to stay one step ahead of their pursuers because they just need to defeat one device/person as opposed to protectors needing to shield everyone, cyberattacks continue to persist.

Will AI help? Maybe, but is throwing more technology with its own cybersecurity challenges at the problem really going to be the answer? We know the answer.

Perhaps we will all just become that much more immune to these cyberattacks that we might feel like they are not as severe or prevalent as years' past, but they will continue, and they will continue to be big.

- Dan Meyer, Executive Editor

Sovereignty will reign supreme

Europe may be making a slight hash of digital sovereignty on the continent, especially as it tries to somehow become an AI powerhouse without American assistance. But it’s worth remembering that there is a worldwide clamor for sovereignty.

Accenture Head of Technology Research for Europe Surya Mukherjee told me in a recent interview that Stanford AI Index research has seen a “meteoric, exponential rise” in incidents of AI in American regulation and legislative documents.

Gartner meanwhile predicts 65% of countries in the world will build sovereignty regulations by 2028, with Mukherjee believing that to be a conservative estimate as he believes that same number has already built up sovereign protections, with Europe’s lead on GDPR now “rubbing off” on the rest of the world.

It’s already become obvious with Cisco, SAP, SUSE, Microsoft, and more investing in EU-compliant solutions within recent months.

I expect to see national sovereign clouds to grow in stature, such as French-focused Orange Business/Capgemini venture Bleu, and SAP’s sovereign Microsoft Azure equivalent for the German market, Delos Cloud.

There will also be more knotty sovereign situations like Canada’s recent demand for French giant OVHcloud to transfer over customer data stored in Europe.

Expect to see more American leaders to tout their cloud sovereign wares next year, following Hewlett Packard Enterprise’s (HPE) lead at its recent European conference. There, CEO Antonio Neri was jumpstarting sovereign AI clouds off the back of it building “pretty much all” of Europe’s supercomputers (thank you, America).

- Giacomo Lee, News Editor

Quantum will get bigger, but won’t be a thing

Market momentum is undeniable, and quantum is on the rise, that’s for sure. What isn’t as certain is when quantum will become a thing, or the aptly named Q-Day, when quantum computers will become powerful enough to break standard public-key encryptions used today.

Just like AI, quantum has its hyperbole, but this year, we’ve seen some genuine breakthroughs: a method to integrate quantum technology into traditional networking systems; a molecular qubit that operates at standard telecom wavelengths; and a “universal translator” to allow disparate quantum systems to communicate over a network with virtually no noise.

Even Nvidia got in on the act with NVQLink, a way for quantum computers to interconnect with classical systems.

But to err on the side of caution with quantum is necessary, given the scale of what it could bring to the world. Nvidia’s Tim Costa, who serves as the company's GM for industrial and computational engineering and quantum, told SDxCentral in October that he doesn’t like to say when quantum computing will become a thing to not set unnecessary expectations.

Quantum will be a thing, but to rush something so inherently complex will only prove disastrous. Unlike AI, where everyone and their dog seems to be developing something, quantum is slowly taking shape and shouldn’t be dismissed so quickly simply because we’ve been talking about it for a while and nothing truly tangible has taken shape yet. So in the words of the guy who fixes Woody in Toy Story 2: “You can’t rush art.”

- Ben Wodecki, Senior Editor

AI gets real

AI was the buzzword for 2025, and has significant momentum heading into the new year. But, as with all things “buzzy,” reality eventually culls an edge.

That will happen in 2026, as the hype around all-things AI begins to turn toward actual use cases. This means that everyone hyping AI as a cure-all will actually move toward real-world use cases that show a real return on investment.

This means not just claims of AI being able to increase productivity or slashing costs, but in being able to provide a sustainable platform that can support cost objectives in real-world scenarios.

This is not to say that marketing teams are going to somehow stop using the term “AI” in every new product or service launch. That’s just crazy talk.

But, I have noticed an increasing number of technical people refraining from such mentions and instead trying to explain new innovations under the guide of actually helping solve problems and not just relying on tech terms to garner interest.

I am also separating this prediction from what many are forecasting will be a full-on bursting of the AI bubble and instead view this moderating view more as a necessary rationalization of AI capabilities and need for the betterment of the broader industry.

- Dan Meyer, Executive Editor

Keep an eye on South Korea

People were worried a laser crunch would stymie enterprise AI goals in the near-future.

Turns out memory was where the real AI shortfall would occur, with South Korean giant Samsung reportedly raising prices of its dynamic random access memory (DRAM) wares by 60% as demand supersedes supply in recent months.

Morgan Stanley meanwhile has noted dynamic random access memory (DRAM) and NAND flash memory prices spike between 50% and 300% in the last six months of 2024.

The result is Samsung Electronics is likely to reclaim its status as the world’s leading DRAM provider from long-term rival SK Hynix. Sources claimed in early December that the division is to account for approximately $10.2 billion of Samsung’s projected $12.2 billion operating profit across the current quarter, up by 422% year on year.

As such, South Korea will stand out as a lynchpin in the AI memory supply chain next year with its dual front of Samsung and Hynix.

And while we’re on the subject of South Korea, AI, and SK, I’d wager that SK Telecom (SKT) will be emulated by AI-evangelizing operators in next year’s AI endeavors.

While SK Group’s telecom arm was dogged all year by a long-running narrative over a major security breach, in the background the firm has been building out its AI strategy with aplomb.

It has effectively, and rather uniquely split into two for all its efforts in the AI space, with one arm on all things telecom, and the other on all things AI, such as its Haein Cluster venture, an AI cloud it’s building for Korea.

With memory at its fingertips, as well as new foundries aplenty from its parent group, SKT will likely become a guiding light in the telecom space. As long as that AI bubble doesn’t burst, that is.

- Giacomo Lee, News Editor

Component shortage, here we come

To go one step further, there’s a genuine concern about components having a knock-on effect on the wider AI market. As detailed above, memory shortages have already begun to rock the industry – Hewlett Packard (HP) is already citing supply chain issues as a genuine roadblock set to impact its profitability in 2025.

Now of course, the firm has been struggling for some time, but to see component issues already being cited by a big-name firm as a genuine problem for 2026 speaks volumes about what's to come. But it’s not just HP: Dell and Lenovo have both warned of cost increases for memory and storage components and are mulling price increases. Apple CFO Kevan Parekh suggested in the firm’s Q4 financial call that it’s seen a “slight tailwind” on memory and storage prices. And Micron has even closed up shop entirely on its consumer division in a bid to focus solely on enterprise AI demand.

Businesses smell such issues like a shark smells blood, and likely won’t hesitate to raise prices or even go as far as to use instances like the looming storage and memory shortage as excuses to cut staff. Of course, the wider semiconductor space has routinely seen peaks and troughs in terms of supply chain issues.

The pandemic-induced shortage was one of several instances of component conundrums, like the 1986 American-Japanese semiconductor trade pact that severely hampered production. But the incident in 2020 exposed the real fragility of the semiconductor space, and prices soared as a result. Another such incident, while not quite at such a scale, could well see similar lead time levels return, and in turn, drag the already fragile tech market down with it.

- Ben Wodecki, Senior Editor

Buckle up for 6G!

We are still a good half-decade away from true 6G networks being commercially deployed, which is just about the right amount of time to engage “ludicrous speed” on 6G hype.

The past few months have already seen some hit “light speed” on their hype, with some talk beginning to nudge toward “ridiculous speed.” But 2026 will require everyone to “fasten all seatbelts, seal all entrances and exits, close all shops in the mall, cancel the three-ring circus, and secure all animals in the zoo.”

This prediction is not without precedent, as we witnessed a similar rush 10 years ago. But, with 5G still viewed as being a financial disaster, there is outrageous pressure for 6G to bring home the cash. And that will drive marketing teams to engage all efforts toward that process.

“You’d better buckle up!”

- Dan Meyer, Executive Editor