SAP Advanced Planning and Optimization, the system IBP was built to replace, will reach the end of mainstream maintenance on the 31st of December 2027, a date SAP has already pushed back four times since the original 2017 target.
That is part of the same Business Suite 7 maintenance window as SAP ECC, which means the two migrations clients have been putting off for years now share the same deadline.
Extended maintenance runs through to 2030, but only for clients who have already committed to S/4HANA by 2027, so the option to simply pay more and wait no longer exists.
The DSAG Investment Report published earlier this year gives a more grounded picture of how this deadline is playing out. Roughly half of the businesses still running ECC plan to convert only by 2030, accepting the higher maintenance costs that come with that delay. A little over a third are targeting 2027 itself. A small remainder are stretching the migration out to 2033.
The APO-to-IBP transition is a wave that will keep generating implementation, integration, and migration work well into the next decade.
The live question is whether AI changes how you should be spending your training time and your project selection over the next two or three years, given that the underlying demand for your skills was already secured by a maintenance deadline that has nothing to do with Joule, agents, or any of the language currently circulating around SAP’s AI strategy.
What Business AI means for IBP
SAP now describes itself as a business AI company, built around a Business AI Platform, an Autonomous Suite of applications, and a new way of interacting with the system called Joule Work
AI assistants, agents, business data, and governance are meant to work together across core enterprise workflows, rather than as a single feature bolted onto one application.
IBP is one consumer of a shared platform layer that also serves finance, HR, and procurement, drawing on the same data foundation and the same conversational assistant underneath all of them. IBP’s AI capabilities will always move at the pace of that shared platform, not at a pace IBP sets for itself.
When SAP improves how its Business Data Cloud handles master data, or how Joule grounds its answers in a client’s actual transactional history, IBP benefits from that improvement automatically, but it also waits for it.
Certain IBP features that sound similar in scope go generally available months apart, simply because they depend on different parts of the platform reaching maturity at different times.
What’s confirmed, dated, and IBP-specific right now
Strip away the marketing language and a clearer picture emerges.
Joule can run master data health checks inside IBP using plain language, revealing inconsistencies in the Manage Master Data app rather than requiring a planner to hunt for them manually. After a planning run, it can explain why demand wasn’t fully met, identify which constraints were binding, and suggest options for resolving the shortfall, comparing two runs side by side if asked.
Forecast explanation has been extended this year to cover the influence of independent variables, offsets, and the features the system selected automatically when building its model. In the Excel add-in, a planner can describe a formula in ordinary language and have the system generate the correct syntax, a feature that reached general availability earlier this year following a customer beta period.
SAP’s own release documentation credits one of these features, the explanation behind recommended safety stock and reorder points, with cutting the time planners spend analysing inventory runs by 30 percent, a figure worth treating as a vendor claim rather than an independently verified result.
These features stop short of certain things, too. The agents capable of autonomously releasing production orders or managing supplier onboarding run inside SAP’s Cloud ERP and Business Network applications, using IBP as a source of data rather than as the place where the work happens. None of this is available if your client is still running IBP on-premise, since every AI capability described here depends on the cloud version of the platform.
Where the hype and adoption don’t match
None of this means the technology is being used at the rate the marketing implies. Gartner’s most recent survey of IT and technology executives found that only 17 percent of organisations have actually deployed an AI agent of any kind, while more than 60 percent expect to do so within two years. Gartner places agentic AI at what it calls the Peak of Inflated Expectations on its own hype cycle, a label that exists to describe the gap between announcement and operational reality.
Other recent industry data finds that, across enterprises generally, only around 31 percent have an AI agent running in actual production rather than a pilot, and a striking 88 percent of agent pilots never make it past the pilot stage at all, most commonly because organisations cannot evaluate whether the agent’s output is reliable enough to trust without close supervision.
That 88 percent failure rate is worth sitting with because it isn’t really a story about AI being immature. The agents stall at the pilot stage almost always because nobody trusts the output enough to remove human supervision, and trust in an agent’s output is just trust in the data and process it was built on, restated in a more visible form.
That’s the trigger worth watching for. The moment evaluation and governance tooling matures enough to let a reasonable proportion of those pilots graduate.
Master data work is the actual AI skill
The evidence so far points to a conclusion that’s easy to miss if you’re only paying attention to vendor announcements.
AI inside IBP depends first on the quality of the data underneath it, and that dependency is not a minor technical footnote. As one SAP implementation partner working specifically on IBP data readiness put it: IBP doesn’t clean up a client’s data, it amplifies whatever is already there, so inconsistent product hierarchies or duplicate customer records produce unreliable forecasts and flawed supply plans regardless of how sophisticated the planning engine behind them is. Planners notice quickly when the output doesn’t match reality, and once they stop trusting the tool, adoption stalls for everyone, AI features included.
This connects directly to the work SAP itself describes as a prerequisite for any of its AI ambitions: the clean core initiative.
Data quality, master data governance, and shared business definitions across departments are not back-office housekeeping, they are AI readiness requirements.
The consulting skill that makes AI inside IBP actually useful to a client is not fluency with a chat interface or familiarity with prompt phrasing. It is the same configuration discipline, the same care with planning area design, and the same insistence on clean, governed master data that has always separated a competent IBP implementation from a troubled one.
There is no separate AI skillset waiting to be bolted onto your existing practice. The skillset you already have, is the AI skillset.
Data discipline is the foundation Joule depends on, not the whole of what AI readiness requires. A client moving an agent from pilot to genuine production still needs the integration architecture, the BTP entitlements, and the governance and evaluation tooling that decide whether anyone trusts the output enough to act on it without checking.
Master data quality determines whether an AI feature produces something true. It doesn’t, by itself, decide whether a business is ready to let that feature act unsupervised.
A consultant who has built strong governance habits has solved the harder and more durable half of that problem, but it is a foundation to build on, not a finished answer.
A consultant whose planning area design is sound and whose master data is clean has very little to fear from any of this, AI or no AI.
The risk this technology introduces isn’t to consultants generally. It’s to consultants whose value was always resting on the gaps AI is now starting to make visible.
Do you need to enrol on a course?
The direct answer, for the great majority of working IBP consultants and S/4HANA consultants who regularly interact with IBP, is no, not yet.
The combination of thin real-world deployment, the gap between what’s announced and what clients have actually switched on, and the fact that the valuable underlying skill is one you already practise, means that formal SAP AI certification isn’t currently earning its place against the other things you could be doing with the same hours.
Become familiar with the foundational requirements for AI to work well: know which features are live, know roughly when the newer ones are scheduled to mature, and know enough about how Joule depends on clean data to confidently advise a client when they ask whether they’re ready for it.
There’s a smaller group for whom that calculation is slightly different.
Consultants working inside firms that have already built structured AI tooling into their delivery methodology, or bidding into engagements where activating an AI assistant has become a contractual line item rather than an optional extra, need to move faster on building real familiarity, because their next project will ask it of them.
That’s a feature of the specific contract or employer, though, not a signal that the wider market has changed.
The four things worth doing
Put the saved time somewhere with a clearer return. Keep deepening the integration work between IBP and S/4HANA specifically, since that’s the skill combination that is demanded more consistently in the job postings, far more often than any AI requirement does.
Read the release notes each cycle, which takes perhaps fifteen minutes and tells you what’s new rather than what’s been restated in a different wrapper. Treat master data governance and clean core literacy as the differentiator worth naming on a CV or in a client conversation, since it’s already billable work and it happens to be exactly what AI inside IBP depends on to function.
And set yourself a fixed point to revisit this whole assessment because adoption data from one point in the year can go stale faster than a feature roadmap does.
Back to the deadline
The 2027 maintenance date that opened this article will define your workload over the next several years. That deadline is what fills project pipelines, drives recruitment, and decides which clients are calling consultancies looking for help.
Awareness of what AI can and can’t yet do inside IBP is worth considering, not because the market is currently paying a premium for it, but because it’s cheap to keep current and it protects you from being caught without an answer the day a client finally asks.
For now, that’s the right scale to give it: present in your thinking, modest in your training budget, and not a primary consideration in how you plan the next stage of your career. This reflects the position on SAP IBP and AI as of late June 2026, and given the pace of SAP’s release cycle, the specific dates above will likely move.
If you are an SAP professional looking for a new role in the SAP ecosystem our team of dedicated recruitment consultants can match you with your ideal employer and negotiate a competitive compensation package for your extremely valuable skills, so join our exclusive community at IgniteSAP.