ERP is no longer just a system of record — it's becoming a system of action, and the shift is happening faster than most procurement cycles can keep up with. For enterprise IT leaders in Saudi Arabia, the UAE, and across the Gulf, this transition is arriving at the exact moment digital transformation mandates and AI investment targets are converging, making agentic ERP one of the most consequential buying decisions of the next two years.

From Copilots to Coworkers: What Changed at SAP and Oracle

The pace of change from the two largest ERP vendors has been striking. SAP's Joule, launched in 2023 as a conversational assistant, has evolved into something categorically different. By Q1 2026, Joule crossed the line from AI assistant to agentic platform, with 40+ specialised agents, 2,400 skills, and Joule Studio now generally available, enabling SAP enterprises to deploy AI that doesn't just recommend — it executes.

Crucially, SAP has built in role-based orchestration rather than one-size-fits-all agents. SAP introduced role-based AI assistants within Joule, each tailored to a specific user role and automatically tapping into the right Joule Agents for the job — so a finance manager forecasting cash flow or an HR professional addressing headcount gets the right insights and actions without guessing which agent to invoke. The bigger structural move is cross-departmental coordination: collaborative AI agents specialized in distinct business functions let Joule execute end-to-end processes, linking actions across departments so finance, procurement, and supply chain deliver measurable results together — for instance, resolving a payment dispute might involve multi-agent teamwork between collections, invoicing, and customer support agents.

Oracle has taken a parallel but distinct path. Oracle AI Agent Studio, announced at Oracle CloudWorld 2025, is the build-and-deploy surface for AI agents inside Oracle Cloud HCM, ERP, CX, and SCM. Rather than a single conversational layer, Oracle shipped a large catalog of pre-built, extensible agents spanning finance, HR, and supply chain — designed for enterprises to configure rather than build from scratch. Importantly, both vendors are converging on the same underlying reasoning infrastructure: both sides of the market now converge on the same underlying model layer, with SAP Joule running on Claude as its primary reasoning engine.

For buyers, the key nuance is governance. Most agentic actions today still sit inside human-in-the-loop guardrails for anything consequential — a detail regional CIOs should press vendors on, since marketing language around "autonomous" agents often outpaces what ships in production-grade, audited configurations.

Why This Is Landing Differently in the GCC

The GCC isn't merely a fast follower on agentic ERP — regional data suggests it may be ahead of global averages on adoption intent. In the Gulf, recent findings show that 19% of organizations have already moved from pilots to full-scale implementation of Agentic AI, with 74% planning adoption. Confidence in ROI is notably high in Saudi Arabia specifically: the Saudi Arabia Tech Report 2026 finds that 76% of Saudi organisations expect AI to deliver measurable ROI at enterprise scale within twelve months — the highest confidence level of any region surveyed, compared to 60% for the UAE. Saudi organizations are also backing that confidence with capital: 39% are committing between $100 million and $249.9 million annually to digital technologies.

The UAE, meanwhile, leads on depth of integration rather than raw optimism. This makes the UAE the leading region for agentic AI integration in KPMG's global survey of 2,500 technology leaders, with 69% of UAE organisations identifying as fast followers — a deliberate strategy to deploy proven technologies at scale rather than experiment at the frontier. This "fast-follower" posture is exactly why SAP and Oracle's agentic rollouts matter so much here: large Saudi and Emirati enterprises are watching global reference customers closely before committing budget, and vendor agent-marketplace maturity is becoming a genuine differentiator in RFPs.

Regional AI adoption overall is also accelerating faster than global benchmarks. AI adoption across the GCC jumped from 62% to 84% between 2023 and 2025, with the UAE reporting 97% adoption across government sectors, while Saudi Arabia has committed $14.9 billion to AI infrastructure in a single policy cycle. But ambition is running ahead of infrastructure readiness in many cases: the conversation has shifted from generative AI experimentation to agentic AI deployment, yet the gap between enterprise AI ambition and production-grade integration has never been wider, for structural reasons. Legacy ERP architecture, fragmented master data, and unresolved data-governance questions are the real bottlenecks — not vendor roadmap slides.

Data Residency and Governance: The Non-Negotiables

Any agentic ERP conversation in the Gulf has to start with sovereignty. AI ambitions now collide with legacy architecture, fragmented data environments, and sovereignty obligations, and Saudi Arabia's SDAIA data residency framework is becoming a regulatory baseline. In parallel, UAE AI governance requirements are advancing, and ISO/IEC 42001 is increasingly influencing enterprise procurement decisions.

This has direct implications for how agentic ERP modules should be evaluated. Before signing any agreement for SAP Joule agents, Oracle AI Agent Studio, or a comparable platform, GCC enterprise IT leaders should be asking:

  • Where does agent reasoning and data actually reside — in-region cloud, or routed through a global model endpoint that may sit outside SDAIA or UAE data-residency boundaries?
  • What level of human confirmation is required before an agent executes a consequential financial, HR, or procurement action, and can that threshold be configured per business unit?
  • How is agent decision-making logged and audited for compliance with emerging governance frameworks, including ISO/IEC 42001 alignment?
  • What is the true cost model — is agent capacity bundled into the existing Fusion or SAP subscription, or does high-volume usage trigger a separate consumption-based fee that could balloon at scale?
  • How portable is the investment — does the agent framework rely on open interoperability standards, or does it lock workflows into a single vendor's orchestration layer?

These questions matter more in the GCC than almost anywhere else, because sovereign cloud requirements, bilingual (Arabic/English) process needs, and public-sector-adjacent compliance expectations mean a generic global agent deployment rarely transfers cleanly into a Saudi or Emirati environment without local calibration.

Getting Agentic ERP Right Requires a Regional Partner, Not Just a Global License

SAP and Oracle's agent rollouts are genuinely powerful — but a licence alone doesn't deliver autonomous finance close, self-service HCM, or agent-orchestrated procurement. It takes a partner who understands Odoo-adjacent and mainstream ERP architectures, GCC data-residency rules, and the operational realities of Saudi and UAE enterprises to configure these agents responsibly and cost-effectively.

Grey Space Computing, an Odoo Ready Partner serving enterprises across Saudi Arabia, the UAE, and India, helps organizations cut through vendor marketing to build ERP roadmaps — agentic or otherwise — that are governed, compliant, and genuinely fit for regional operations. If your organization is evaluating SAP, Oracle, or alternative ERP paths for the agentic AI era, talk to GSC's ERP advisory team about a readiness assessment before you commit budget.