Results and prospects for AI in business applications in 2026

AI hype cycle as internet advent 2.0

However, settling the cyber security problems that could yet bedevil enterprise applications is more tactical in nature than it is strategic for businesses and other organisations.

The question of how the AI hype that was dramatically boosted with the advent of generative AI in 2022 – and the maturation of that into agentic AI in 2025 – can be productively compared and contrasted with the rise of the internet in the late 1990s is one that commands more strategic consideration. It is also a plausible guide to what happens next, in 2026.

In a strong sense, we have been here before. Answers to that question of comparison can suggest likely lines of development for AI in enterprises in 2026. Computer Weekly posed the question of comparing these two phenomena in technology history to a series of leading IT industry executives in 2025, during the fourth-quarter conference season.

Steve Miranda, executive vice-president of Oracle applications development at Oracle said, ahead of Oracle AI World, that he would not pick out customer experience (CX) as the most likely area for AI development: “Around 1998 or thereabouts, we at Oracle had a slogan – ‘the internet changes everything’. If you recall at that time, there was a ton of focus on e-commerce.

“Well, e-commerce certainly was there, but the internet changed music and it changed the way you listen to music. It changed the way you watch – I won’t even say watch television, as such. It changed telecommunications and meetings. It changed everything.

“It’s our belief that right now that AI will change everything. There are charismatic use cases right now, like customer service and CX. But when we say it’s changing everything – and I’m giving examples of invoice scanning and payments and dealing with [a company’s] ledger – I don’t think anybody else appreciates the fullness of the impact this is going to have across the board. We’re at the same stage as with the internet now, except it’s going faster.”

At SAP TechEd in Berlin, Philipp Herzig, CTO of SAP, said that how enterprise applications develop will be more interesting than how infrastructure and the large language models (LLMs) themselves develop. The current rise of AI is, he said, going much faster than the internet, adding: “But I still observe that it is very comparable to the beginning of the internet. There was also a lot of focus on the infrastructure plans of the companies that built the switches and the routers and so on, back then.

“Then the discussion moved to ISPs [internet service providers] – the AOLs and the telecoms [companies] that brought the internet to the home and built distribution to consumers. But, really, the value accrued in cloud services that we are all consuming today in streaming services and cloud software.

“Our conviction is that even though there’s a lot of discussion occurring on the hardware side and around datacentres and GPU capacity, for example, this is not where the end game will be played. The end game will be played in amazing user experiences and AI-native software as a service that people can just turn on. We will see a lot of new business models that will emerge out of that.”

That was in Berlin, in early November. In Barcelona, later that same month, Gerrit Kazmaier, president of products and technologies at Workday, said at Workday Rising EMEA: “The key thing to understand is that in technology, you have cycles, and the first piece that always happens is the infrastructure innovation. You saw this with Broadcom, and so on, with broadband internet. The cloud infrastructure that was built on the availability of bandwidth and the economics that allowed you to do business over theiInternet ultimately created SaaS [software as a service].

“We’re seeing the same cycle. Now we are in the second stage. We had the creation of very powerful computer infrastructure. Now we have the form factor of large language models for basically utilising that to create a different type of computation – you could call it computational reasoning.

“That is something that didn’t exist before, and now software vendors that basically solve business problems have the ability to take that and change the way business systems are built. Business processes to a large extent are designed around the constraints in human reasoning…But now with computational reasoning, we have the chance to completely rethink.”

At the end of the year, David Richardson, vice-president of AgentCore AWS, told Computer Weekly at ReInvent in Las Vegas: “I have long-term optimism, which is important because there will be bubbles and things that happen, ups and downs. The beauty of the internet was that it made it possible to have incredibly flexible communication amongst a wide variety of different compute capabilities, and what people built on top of that nobody would have ever predicted.

“I first used the internet in 1987, seven years before I ever used the Web, and it was already fascinating. The other thing that happened with the internet is that it created a massive ecosystem about building up the internet…[And now] there are so many different participants in the ecosystem who see the opportunity to deliver value through innovation that it does feel like we have many years ahead of us of increasing reasoning and resiliency capabilities out of these models.

“We will keep finding new ways to apply those to problems that we either thought we couldn’t address before or couldn’t address economically. So, that’s my many-year view as to what will happen – 12 months from now, I don’t have a clue and wouldn’t make a forecast about that.”

What does this fundamental comparison of the AI hype cycle and the dawn of the internet at the turn of the twentieth century tell us about the likely path of development of AI-infused business applications in 2026, and beyond? It tells us that a new stage is ready to begin. Classical AI – so, machine learning, to be simple – has now joined generative AI and agentic AI in ways that will mean enterprise applications will function more like video streaming services that algorithmically deliver content to consumers, and less like the old broadcast television that Generation X grew up with.

Claus Jepsen, CTO of Unit4, suggestively calls this “ambient ERP”, and his predictions for AI in ERP in 2026 make interesting reading. Similarly, Steven Webb, the UK chief technology and innovation officer at Capgemini, makes a convincing argument, in a 2026 predictions opinion piece for Computer Weekly, that “AI is no longer a supporting technology. It is rapidly becoming the operational fabric of modern enterprises. We’re seeing a decisive shift beyond single-task activities towards autonomous, adaptive and self-optimising systems powered by multi-agent systems.”

That level of autonomy will require technologies and processes of governance that have been developing since late 2022, when generative AI first burst on to the scene, but which will need to catch up.  Michelle Eisenberg, general counsel at Unit4, has interesting points to make about building an AI governance framework, balancing innovation with circumspection that go beyond its origins at that supplier.

But make no mistake, as Brad Novak from DXC Technology – who has been through the “mobile, social, cloud and big data” wave, and the rest – told Computer Weekly at the 2025 Boomi World Tour in London: “This is different, it is as if we all woke up with another 20 IQ points.”