Businesses will get unlimited agentic AI pricing, data tools will be a headache, AI agents will be more like a feature than a revolution, and physical AI will be big. These are some of the trends to watch out for in 2026.
Here’s a look at business technology trends and forecasts for 2026, grouped by confidence levels.
High trust
Agent Business Licensing Agreements Will Become the Norm as Executives Move Against Them. In 2025, enterprise software publishers, worried about losing places, introduced consumption models. The idea was that a hybrid approach would be the best of both worlds for CIOs. However, consumption patterns were unpredictable and CIOs, not to mention CFOs, wanted predictability. Enter the Agentic AI Enterprise License.
Miquel Milano, President and Chief Revenue Officer of Salesforce, discussed the rationale behind Salesforce’s Agentic Enterprise License Agreement (AELA). “AELA is for customers who have already experimented. They’re ready to scale. They want to go all out, so we agree on a flat rate, and then it’s a shared risk,” Milano said.
ALEA is all you can eat with Agentforce or any other integrated cloud. Of course, once you’re all in Salesforce, you can monetize the platform better in the next deal. In 2026, it is very likely that you will see similar arrangements from SaaS providers.
Here’s the problem: SaaS providers can sign FTAs at a loss when playing for renewal (when you’re completely locked in). Milano looks at the lifetime value of a customer. “If the customers are smart, they can rob the bank. They can really make a lot of money out of it. We take the risk because we want our customers to succeed. I’d like to have a customer where I price an AELA at $5 million more, and the customer has deployed so much that the deal is not profitable for me. If the deal is not profitable for me, that means the customer is the happiest customer in the world. And then I have another 20 years to monetize that customer,” Milan says.
Enterprise data costs and API economics are going to be a headache. Celonis sues SAP for access to data. In October, U.S. District Judge Vince Chhabria in San Francisco ruled that SAP should face the Celonis lawsuit. Towards the end of 2025, The Information reported that Salesforce increased the prices of applications that exploited its data. CIO.com noted that these connector costs are likely to impact IT budgets.
As agentic AI is deployed and agents come online, there will be multiple skirmishes in 2026 over these data tolls. There will be a standard toll to cover computing costs, but companies must remember that they own the data. In some cases, your provider may think otherwise.
Connection fees will be the new way out of the cloud to move data. I would argue that data costs will be the biggest risk to scaling AI agents.
Agentic AI is a feature and the real deal is decision speed. This theme is Michael Ni’s department, but decision speed (how quickly smaller decision trees and processes can be automated at scale) will be a central theme. When Ni, during a video chat, said that agentic AI was just a feature, I had two reactions. First reaction: how many pixels were spent on a feature?!? Second reaction: Mike is right. The game is about decision speed and your task in 2026 will be to put the pieces in place to make it happen.
“What we’re seeing is that the first decisions are smaller decisions, with automations on the back leading to human engagement. Then we start to shrink those decision trees to get some of the 5x, 10x improvements. We’re seeing leaders actually able to realize results through decision automation,” said Ni, who noted that 2025 was a year where many of the building blocks of decision velocity were put in place.
Advanced internal deployment is becoming a necessity. Software companies – all taking inspiration from a Palantir play – talked about forward-deployed engineers almost every time executives talked about agentic AI deployments. In 2026, companies will realize that they need their own forward-deployed engineers to work on data, process, architecture and AI automation. These engineers will know the business and industry better than those you borrow from software companies and service companies.
Average confidence
The AI market will bifurcate as the bubble bursts in 2026 or 2027.. Concerns about capital spending and debt in AI infrastructure will be what really increases in 2026. But don’t get distracted by your AI plans. There is a good chance that 2026 will highlight the end of the easy money needed to form LLMs, raising capital at a ridiculous rate and remaining fabulous performance obligations. This AI market, dominated by the circular economy and OpenAI of AI, will be completely different from the enterprise AI version. This productivity boom is only just beginning, as AI and process automation converge.
Build beats buy. The build versus buy debate continues as AI agents make it easier to build apps you previously bought. Additionally, building seems like a great option since customers are starting to push back against the inflation of SaaS offerings. SaaS costs are increasing every year. It’s almost as bad as health care.
In 2026, there will be an inflection point where businesses are convinced that applications tailored to their use cases are the way forward. Sunk costs in business systems will be eliminated by using agentic AI as the user interface.
The benefits of AI extend to more levels of the business stack. Software companies will see revenue and productivity gains thanks to AI. On Wall Street, enterprise software companies are finally joining the AI rally. Nvidia and AI infrastructure games will be stable in 2026.
Physical AI has its moment of glory. Manufacturing and industrial sectors are starting to leverage physical AI to deliver real value. Additionally, breakthroughs in physics AI will begin to rival the early days of LLMs. This focus on physical AI sparks optimism about robotics as well as cutting-edge AI applications.
Not absurd, but unlikely for 2026
- Nvidia shares close the year flat or down, but the drop isn’t enough to drive real selling. Sales growth is starting to slow as hyperscalers rely more and more on their custom silicon. The industry is working to break Nvidia’s hardware and software divide around AI.
- AI infrastructure is overbuilt due to hardware and software advancements that require less computation and energy. Wall Street rewards companies like Apple that haven’t gone crazy building AI infrastructure. AI backlash is growing as multiple local NIMBY efforts thwart plans to build data centers in small towns and rural areas. Note that I said something similar in 2025 to no avail.
- OpenAI realizes it can’t raise money continuously and indefinitely continues its austerity campaign to show that it can generate cash flow and become profitable.
- Quantum use cases are becoming more widespread across the enterprise as quantum supremacy arrives early. It’s also becoming clear that superconducting quantum computing is the clear technology winner and that reality is causing a mad rush for companies focused on trapped ions, neutral atoms, annealing and other techniques.
- Meta is reorganizing its AI operations again once it’s clear its new management team and direction aren’t delivering results. Meta’s AI operations are like the New York Mets, a massive payroll that doesn’t generate wins.
- High memory costs are creating a strike by buyers of PCs, servers and smartphones.
- 2026 is becoming one of the most important years ever. for IPOs like Databricks, Anthropic, OpenAI, SpaceX and Stripe all go public. Two of these five headliners are trading below their IPO prices after 3 months.
- TikTok usage plunges under new ownership as US algorithm is tweaked.
Dashboard for 2025
Here’s a preview my predictions for 2025 it worked and other areas where my crystal ball was cracked. In 2025, I included the probability of a prediction coming true.
On target:
- The year 2025 has been more volatile than usual.
- Productivity gains from AI have expanded across businesses.
- OpenAI and Microsoft have undoubtedly become enemies.
- Companies will try new SaaS revenue models and scream. Consumption seems great…until you get the bill.
- ERP is under fire from critics. ERP hasn’t gone anywhere but is being abstracted.
Off base:
- This supposed attempt at supplier consolidation never happened.
- Agentic AI will usher in autonomous processes. We are clearly only in the early stages. The buy side is still wary of foreclosure.
- AI data center construction will stagnate. In fact, it accelerated a few days after I made this prediction.
- Nvidia’s growth is slowing. Nvidia still faces the rule of numbers and increased competition, but 62% revenue growth is pretty sweet.
- Edge computing is becoming more critical for AI workloads. It will happen, but it certainly won’t happen in 2025.