Image credit: Getty Images
As 2025 draws to a close, global enthusiasm for artificial intelligence is increasingly tempered by a rising wave of doubt. Stunning AI breakthroughs now sit uncomfortably beside warnings of an overheated market, an overbuilt data-center ecosystem, energy constraints, and a growing list of GenAI pilots sputtering out before reaching meaningful business value.
SAS experts and industry leaders say 2026 is shaping up to be a defining year, one in which AI’s power brokers and enterprise adopters will be forced to justify return on investment, implement credible governance, and confront complex technical and ethical questions that can no longer be postponed.
Despite looming concerns, optimism endures. SAS thought leaders underscore a unifying message: the future of AI depends on accountability. The organisations that embrace strong data foundations, responsible governance, and strategic clarity are best positioned to ensure AI matures into a truly transformative force, one that enhances human capability, fuels organisational performance, and accelerates innovation across industries.
Read more-From Dubai to Riyadh: Could AI be your next workplace colleague?
After years of frenetic buildouts, 2026 will expose the economic imbalance behind sprawling data-center expansion. Despite soaring expectations, many projects fail to produce revenue that covers the immense capital and operational costs. As investors and operators reassess their bets, technology providers will be forced to seek new architectural and financial models.
“Major investments in data center buildouts will prove impractical as costs come home to roost,” says Jared Peterson, SVP, Platform Engineering. “Expect economics experts to issue pointed reminders that they warned this outcome was inevitable.”
The combination of infrastructure strain, rising energy prices, global regulatory scrutiny, and persistent supply-chain bottlenecks suggests that next-generation AI architectures will need to move beyond brute-force scale. The question is no longer how fast data centers can expand, but how efficiently AI workloads can run.
The AI spending shake-up
After a two-year mania in which companies poured billions into speculative AI tools, wrappers around ChatGPT-style models, and unproven productivity claims, CFOs are drawing a firm line.
“After billions wasted on ChatGPT wrappers and vaporware, CFOs are demanding real ROI,” says Manisha Khanna, senior product manager, AI & Generative AI. “The honeymoon phase where ‘AI innovation’ justified any budget is over.”
With cost-per-query, model accuracy, and measurable business outcomes under the microscope, generative AI projects will need to show tangible progress within six to 12 months or risk being mothballed. Vendors unable to demonstrate material value will face a growing churn cycle as enterprises reallocate budgets toward practical, outcome-driven solutions.
CIOs step into a new role: Chief Integration Officers
As agentic AI systems proliferate across organizations, the role of the CIO will expand dramatically. No longer primarily technology enablers, CIOs will evolve into orchestrators of complex, multi-agent ecosystems.
“In 2026, CIOs will answer the call to orchestrate the agentic AI future,” says Jay Upchurch, CIO. “The role decisively shifts from tech enabler to ecosystem integrator.”
The growing patchwork of autonomous agents, foundational models, enterprise platforms, and regulatory expectations requires new forms of coordination. Integration, governance, cross-functional alignment, and architecture design will dominate the CIO agenda as agent-led systems reshape IT operations.
With SAS marking half a century of innovation, 2026 will usher in a new reality, one in which AI agents transition from tools to true collaborators.
“Enterprises are evolving into ecosystems where AI agents are no longer tools; they are teammates,” says Udo Sglavo, VP, Applied AI and Modeling R&D.
In these mixed human-AI environments, agents will execute tasks, share context, and continuously learn alongside employees. This shift challenges long-standing norms about decision-making, productivity metrics, and organisational design.
Agentic AI: Accountability for profit and loss
Agentic systems will not only advise; they will increasingly operate autonomously. According to Iain Brown, head of AI & Data Science, Northern Europe, by the end of 2026 Fortune 500 companies will report that AI agents autonomously handle more than a quarter of multi-step customer interactions.
“These agents won’t just advise, they’ll execute, with measurable revenue impact,” Brown says.
This shift will spur new roles such as Agent SRE and even the chief agent officer. But it also introduces new risks. The first major “agent outage,” Brown predicts, will make headlines as enterprises learn that when AI systems drive revenue, downtime becomes financially, and reputationally, significant.
With the debate over workforce disruption reaching fever pitch, 2026 will mark a pivot toward AI-enabled empowerment.
“It is becoming increasingly clear that AI should empower people, not replace them,” says Bryan Harris, CTO. “Leaders must invest boldly in their workforce through continued change.”
Companies that prioritise augmentation over automation will be better equipped to compete, innovate, and attract talent in a tightening labor market.
Organisations that raced into AI adoption without addressing governance, risk, and Responsible AI principles may face an uncomfortable 2026.
“Mature, early AI adopters that bypassed attempts to measure and incorporate AI responsibly will be exposed,” warns Luis Flynn, market strategist for Applied AI, Open Source Software & ModelOPS. He predicts a “massive loss of credibility” once the use of commoditised, poorly governed AI systems becomes public.
The comparison to the log4J breach underscores the scale of potential fallout.
The longstanding debate pitting innovation against trust will dissolve in 2026, replaced by the recognition that governance is not a brake on progress but a foundation for it.
“As government regulation remains inconsistent, corporate self-governance will extend to include the necessary guardrails,” says Reggie Townsend, VP, Data Ethics Practice.
“Governance isn’t a restraint on innovation, it’s a necessary companion.”
Organisations that invest early in governance frameworks, explainability, and oversight will pull ahead of faster-moving but less disciplined competitors.
Sovereign and hybrid AI architectures rise
Regulated industries and global enterprises increasingly demand control over their data, models, and infrastructure. “Bring your own model” and sovereign AI configurations, where companies host foundation models within their own governance boundaries, will become the default.
“The cloud stays, but control shifts,” says Marinela Profi, global agentic AI strategy lead.
This architectural realignment reflects a broader trend: enterprises want full stack visibility and accountability while retaining access to scalable cloud resources.
Energy availability is emerging as a critical constraint for AI progress. By 2027, US data centers will require an additional 29 gigawatts of power. A five-gigawatt center running at 90 per cent capacity consumes more annual energy than 3.65 million homes.
Without rapid expansion of national energy infrastructure, the US risks falling behind global competitors.
“Without increasing power capacity as quickly as possible, the US will lose its ability to collaborate and advance in AI and data on the international stage in 2026,” warns Franklin Manchester, Global Insurance Advisor.
In 2026, agentic AI will transition from experimental pilots to the operational core.
“Those investing in the right infrastructure, governance and skills will unlock smarter decisions and seamless experiences,” says Jennifer Chase, CMO. “Those who don’t will fall behind.”
This shift will define competitive advantage across sectors, from customer service to supply chain to financial services.
Quantum computing heats up
The quantum computing market will accelerate significantly in 2026, with expectations that the technology could deliver early-stage commercial value by 2030.
“Investors will broaden their scope from hardware and cryptography to software and applications,” says Amy Stout, Head of Quantum Product Strategy.
Companies will expand hiring for quantum expertise and begin exploring holistic “quantum architecture,” encompassing the full computing stack.
Synthetic data will transition from workaround to competitive differentiator. Limited access to high-quality, multimodal real-world data, combined with tightening privacy requirements, will make synthetic data essential for model training and innovation.
“Expect a data arms race,” says Alyssa Farrell, senior director, Platform and Horizontal Solutions. “Winners will be those who master synthetic realism and shift from experimental to essential capabilities.”
HR’s new mandate: Managing humans and AI agents
As agentic systems become embedded in daily workflows, HR will be responsible for a hybrid workforce.
“HR leaders will manage more than people, they’ll manage AI agents too,” says Jenn Mann, CHRO.
New policies, onboarding processes, and performance frameworks will be required as AI becomes a digital coworker influencing team dynamics and decision-making.
According to Stu Bradley, SVP, Risk, Fraud and Compliance Solutions, 2026 will inaugurate a long-awaited market correction.
“Hype collides with governance, and only accountable innovation endures,” Bradley says.
As oversight intensifies and demand for clear ROI grows, the market will weed out vanity projects and reward disciplined execution. Overhyped technologies will fade; responsibly deployed, operationally rigorous AI will rise.
Across industries, AI is evolving from tool to teammate, reshaping delivery methods, team structures, and definitions of success. From intelligent testing and code generation to warehouse orchestration and predictive logistics, AI is becoming embedded in the core of delivery operations.
A recent workshop series reflected in Deloitte’s 2026 AI series: When AI becomes a team member found that organisations must begin planning today for future AI-enabled delivery teams. The report emphasises that while 2025 conversations focused on productivity and job displacement, a deeper shift is in motion: a fundamental redesign of what constitutes a high-performing team when AI is at the center of its operating model.
Despite an explosion of AI use cases, several foundational questions remain unresolved:
- Team operating models have yet to fully integrate AI, how should hybrid human/AI teams be structured?
- New roles such as prompt engineers, model reviewers, and AI delivery leads remain emergent, how should responsibilities evolve?
- Accountability is unclear when AI tools influence or make decisions, who governs outcomes?
- AI fluency remains low across most teams, how should organisations upskill workers for AI collaboration?
Designing high-performing hybrid teams
Deloitte outlines four key steps to redesigning delivery teams for an AI-augmented future:
- Define AI’s role
Organizations must determine whether AI acts as an assistant, peer reviewer, or specialist. This clarifies human responsibilities, validation processes, and workflow design.
- Create hybrid roles
New positions will bridge human-AI interactions, including:
- AI delivery lead
- Prompt architect
- AI risk steward
- AI quality assurance specialist
- Embed governance
Governance must be built into daily delivery rituals:
model validation in the definition of “done,” AI reviews in retrospectives, documentation of AI logic, and more.
- Upskill for fluency
Success hinges not only on tool familiarity but on cognitive skills such as interpreting AI outputs, managing uncertainty, and knowing when not to use AI.
Where to start
Deloitte recommends practical starting points:
- Pilot AI as a team “member” in one delivery cycle
- Conduct an “AI Role Canvas” exercise
- Add AI validation to CI/CD or release gates
- Evaluate outcomes beyond speed, quality, trust, transparency
The biggest challenges identified in workshop feedback were change management and governance design, areas where traditional tech teams lack experience. Yet with thoughtful design, hybrid teams can achieve dramatic productivity gains.
AI is not replacing delivery teams, it is reshaping them. The organisations that thrive will be those that intentionally redesign how humans and AI collaborate, make decisions, and deliver value. AI is here to stay; the question is whether teams are prepared to work alongside it.
