SAP vs IFS vs Infor: How Each Is Shipping Agentic AI in 2026

Jeet Poptani, Chief Transformation Officer at AumentoAI, compares what's real across the three ERP platforms he delivers transformation programmes on.

Every major ERP vendor is now making the same pitch: agents, not just automation; outcomes, not just data entry. Most comparisons of that pitch come from analysts who've never sat inside a delivery programme for any of these platforms, or from a vendor's own SI partner, whose comparison somehow always favours the platform they resell. I work across SAP, IFS and Infor engagements, which is a narrower vantage point than an analyst's but a more honest one: here's what each vendor has actually shipped as of Sapphire and H1 2026, not what's on the roadmap slide.

SAP: platform consolidation, at platform scale

SAP's Sapphire 2026 announcements unified Business Technology Platform, Business Data Cloud and Business AI into one Business AI Platform, with an SAP Knowledge Graph mapping business entities underneath and Joule Studio as the build environment on top. The Autonomous Suite now spans more than fifty domain agents across finance, supply chain, procurement, HR and customer experience, with SAP's own Autonomous Close Assistant claimed to compress financial close from weeks to days. RISE customers get three Joule assistants activated by default; GROW customers get the full portfolio, backed by a €100 million partner adoption fund.

The honest read: this is the most architecturally ambitious of the three — a genuine platform play, not a bolt-on. It's also the one furthest from your specific process today. Fifty agents is a catalogue, and the gap between catalogue and safe deployment is data readiness and governance, not technology.

IFS: agentic AI on asset and service data, not general ledger data

IFS Loops deploys "Digital Workers" against field dispatch, supplier coordination, order management and inventory replenishment — roughly 50 agentic skills live now, on a roadmap toward 100+, with a Loops Studio release letting customers build their own. IFS reports 60% of agentic transactions on the platform now fully automated end to end, alongside 25% ARR growth and 84% recurring revenue in H1 2026, with enterprise wins including Coca-Cola, China Airlines, Miele and First Solar.

The honest read: IFS's AI is landing on a different, arguably more tractable problem than SAP's — high-volume operational decisions in asset-heavy environments, rather than general enterprise-wide autonomy. The 60% automation figure is a real number about IFS's own customer base, not a guarantee about yours, but the underlying thesis — agentic AI works best on well-defined, high-volume decisions — is sound and worth taking seriously if your business is service- or asset-heavy.

Infor: industry-specific agents, built on someone else's AI stack

Infor's Industry AI Agents target eight core industries and their micro-verticals — dairy production, EV manufacturing, textile fabrication among them — running through an "Infor Agentic Orchestrator" built on AWS with Amazon Bedrock underneath. Infor's customer proof point, State Electric Supply, reports process-issue identification 86% faster, shipment times cut from five days to two, and payment processing from seven days to three.

The honest read: Infor is making a narrower, more industry-vertical bet than either SAP or IFS, and being transparent that it's built on a hyperscaler's AI infrastructure rather than proprietary models is, if anything, a point in its favour for CIOs already standardising on AWS. The State Electric Supply numbers are a single case study, not a platform-wide statistic — treat them the way you'd treat any vendor reference case, as an existence proof rather than an expected outcome.

What's actually consistent across all three

Strip the branding from each vendor's keynote and three things are true everywhere:

• Every platform's agentic AI is only as good as the process and data underneath it. SAP's clean-core requirement, IFS's asset/service data quality, and Infor's industry-specific process maturity are the same problem wearing three different vendor logos.

• No vendor has solved multi-vendor AI governance, because it isn't their problem to solve. If your enterprise runs SAP finance alongside IFS field service or Infor distribution — a common state after any M&A or multi-entity history — you are the one who has to govern Joule, Loops and Infor's agents under one risk framework, one incident response process, and one EU AI Act compliance posture. None of the three vendors' roadmaps addresses this, because each is naturally scoped to its own platform.

• Every vendor's headline statistic is a reference-customer number, not your number. Fifty agents, 60% automation, 86% faster issue resolution — all real, all earned by a specific customer's specific process, and none of them predictive of your result without your own pilot and your own instrumentation.

How I'd actually choose, or combine

The right question isn't "which vendor's AI is best" — it's "which vendor's AI is closest to the specific decisions my business makes most often, at the volume that makes automation worth the governance overhead." A finance-heavy, standardised-process enterprise will get more from SAP's Business AI Platform sooner. An asset-heavy or field-service-heavy operation will likely see IFS Loops pay back faster on a narrower set of workflows. A distribution or manufacturing business already committed to AWS has a real case for Infor's approach.

Most enterprises I work with aren't choosing one — they're running two or three of these platforms simultaneously across different business units, which makes the governance question, not the platform question, the one that actually determines whether 2026's agentic ERP investment pays off or becomes next year's audit finding.

AumentoAI advises CIO, COO and CFO offices on evaluating and governing agentic AI across SAP, IFS and Infor environments — including multi-platform enterprises running more than one. Book a Value Advisory Session to get an honest read on your platform's AI roadmap.