2026 becomes the test for AI’s long-term future

• Companies head into 2026 expecting clear returns from AI, not experiments.
• Agentic AI set to join human teams in 2026.

As 2025 closes, interest in AI remains high, but a degree of unease around a possible bubble, rising energy demands, and early GenAI projects may be falling short of expectations.

Experts at data analysis company SAS say 2026 will be a turning point as boards and senior leaders will expect clear results, tighter oversight, and answers to long-standing questions about ethics, cost, and value.

While some concerns are real, SAS leaders point to a common theme: the next phase of AI depends on accountability. Both providers and enterprise teams will have to return to basic data practices and commit to responsible use if they want long-term value.

Data centre spending under pressure

Heavy investment in new data centres may start to look unrealistic. Many firms poured money into large builds, expecting strong returns that never came. If the income does not cover the cost, companies will search for cheaper models, and economists will call it a predictable outcome.

Budgets shift as spending demands get sharper

CFOs want proof that generative AI programmes are worth the cost. After years of spending on tools that offered little more than a user interface on top of a model, leadership teams now ask about cost per query, accuracy, and whether the work leads to real savings or growth. Projects that cannot show value in half a year may be paused or replaced.

Manisha Khanna, Senior Product Manager, AI & Generative AI, says the early wave of “AI innovation” spending is over, and companies will expect results rather than experimentation.

CIOs move into full-scale integration

As AI agents spread in systems, CIOs will be asked to integrate them into the wider IT structure. This shifts the role from general oversight to full-stack coordination in teams, policies, and architecture.

Jay Upchurch, Chief Information Officer, describes it as the rise of the “Chief Integration Officer,” where governance and cross-team work become daily priorities.

AI agents become part of the workforce

Enterprises will begin treating agentic AI systems as teammates rather than tools. Agents will handle tasks, share context, and adjust as they work with people.

Udo Sglavo, Vice President, Applied AI and Modelling R&D, says companies will operate with mixed teams of humans and agents who learn alongside one another.

By the end of the year, Fortune 500 firms may report that agents resolve a quarter of multi-step customer issues on their own. They will perform work that affects revenue, which could create new jobs like Agent SRE or Chief Agent Officer. But it may also lead to the first large-scale outage tied to an autonomous system.

Dr Iain Brown, Head of AI & Data Science, Northern Europe, notes that once revenue depends on these systems, downtime carries a direct cost.

Human-centred adoption becomes the preferred path

Leaders will need to decide whether AI replaces work or strengthens it. The trend points toward the latter, with a growing push to support employees rather than remove roles.

Bryan Harris, Chief Technology Officer, says companies will need leaders “bold and inspirational” enough to support their people through ongoing change.

Early shortcuts in AI come to light

Some early users rushed ahead without strong testing or governance. In 2026, those shortcuts may surface where companies that relied on weak or poorly understood models could face questions about trust and reliability.

Luis Flynn, Market Strategist for Applied AI, Open Source Software & ModelOPS, warns that this could lead to a “massive loss of credibility.”

Governance becomes a practical requirement

Enterprises will no longer treat trust and innovation as opposing ideas. With government rules still uneven, companies will set their own guardrails. Those that treat governance as part of their strategy, rather than a barrier, are more likely to succeed.

Reggie Townsend, Vice President, Data Ethics Practice, says the firms that do well will be the ones that pair strong oversight with clear plans for use.

Sovereign and hybrid AI gain ground

Firms want more control over their data and models. This may drive a shift toward “bring your own model” setups, where companies run AI in their own compliance boundaries. Cloud platforms remain part of the mix, but control moves inward.

Marinela Profi, Global Agentic AI Strategy Lead, expects these setups to become standard in regulated sectors.

Agentic AI moves into core operations

Agentic AI will no longer sit in pilot programmes. It will become part of everyday work, influencing how firms manage decisions and customer service. Companies with the right systems and skills will gain an edge. Those without them may struggle to keep up.

Jennifer Chase, Chief Marketing Officer and Executive Vice President, says this shift will separate companies that are ready for autonomous systems from those that are not.

Quantum investment expands beyond hardware

Interest in quantum computing will rise through 2026. Investors will widen their focus past hardware and cryptography to include the software and application layers that bring practical value. Demand for staff with quantum skills will grow.

Amy Stout, Head of Quantum Product Strategy, points to the rise of “quantum architecture,” which covers the full stack of a system.

Synthetic data becomes a core competitive area

With real-world data still limited by privacy and access issues, synthetic data moves into the mainstream. Companies will try to build convincing, useful datasets that help them train models safely and at scale.

Alyssa Farrell, Senior Director, Platform and Horizontal Solutions, calls it a “data arms race,” where the advantage goes to those who can produce realistic synthetic data at volume.

HR prepares for a mixed workforce

Human resources teams will be responsible for both people and AI agents. As agents handle more tasks, HR will create new rules for onboarding, reviews and day-to-day cooperation between staff and digital coworkers.

Jenn Mann, Chief Human Resources Officer, says the future workplace will be part human, part machine.

A shift toward accountability

SAS experts expect 2026 to be the year when the gap between hype and real value becomes clear. Oversight, measurable results and honest reporting will decide which projects continue.

Stu Bradley, Senior Vice President, Fraud & Security Intelligence, says this change will close down weak projects and shift investment back toward strong data foundations, reliable models, and clear governance.

As the hype fades, the industry faces two questions: how deep the reset will go, and when the next chapter of steady, accountable growth will begin.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.