SAP Trends 2026: What Sapphire Told Us and What It Means for SAP Customers

Sapphire 2026 confirmed what most SAP customers already suspected: AI isn’t a feature bolted onto SAP anymore, it’s the organising principle of the whole portfolio. Here’s what actually came out of the event, what’s genuinely new versus repackaged, and what it means if you’re running SAP right now.

The SAP Business AI Platform

SAP used Sapphire 2026 to consolidate its AI story under the SAP Business AI Platform (BAIP), pulling Joule, the generative AI hub, Joule Studio, and the underlying data and agent infrastructure together under one commercial and technical umbrella rather than a scattering of separately branded pieces. Practically, this also came with a credential rename, so if your team has been tracking SAP AI certifications under older naming, it’s worth checking those map across correctly under BAIP before you plan next year’s training.

The consolidation matters more than the branding. For the last three years SAP’s AI story has been genuinely useful but genuinely fragmented, Joule here, generative AI hub there, Joule Studio somewhere else again. Putting it under one platform name is SAP’s acknowledgment that customers were struggling to see how the pieces fit together, and a signal that they’re meant to be bought and adopted as a connected whole from here on.

Joule Work and Joule Studio Reach General Availability

Two milestones worth noting for anyone still on the fence about agentic AI in SAP:

Joule Work is now positioned as the everyday, role-based assistant layer, the thing an individual employee actually talks to inside their SAP applications to get a task done.

Joule Studio, the tool for building and configuring custom AI agents, reached general availability at Sapphire 2026. That’s the more significant milestone for most Bluestonex clients. Up to now, building custom agents has involved a fair amount of “early access, subject to change.” GA means it’s now a supported, production-ready part of the stack, not a preview feature you build on with one eye on the roadmap. If custom agent development has been sitting in your “wait and see” pile, that’s the signal to move it into “start planning.”

SAP Knowledge Graph

Also central to the AI story is SAP Knowledge Graph, which maps relationships and business context across SAP’s data landscape and feeds that structured understanding into Joule and the wider BAIP. This is the layer that lets an AI agent understand what a customer, an order, and a shipment actually mean to each other, rather than just seeing rows in separate tables. Without it, agentic AI in SAP is guessing at context. With it, agents can reason across relationships the way a knowledgeable employee would.

The Anthropic Partnership

SAP’s relationship with Anthropic has been building for a while (SAP was among Anthropic’s earliest enterprise investors back in 2023), and it’s now showing up in production: SAP Joule for Consultants uses Anthropic’s Claude models via Amazon Bedrock to give consultants natural-language access to SAP’s enormous certified knowledge base, tens of millions of documents’ worth, with clear, cited answers rather than a documentation search. It’s a good example of what “AI built into SAP” actually looks like in practice, not a chatbot bolted onto the side, but a capability that draws on a specific, authoritative body of SAP knowledge.

RISE and GROW with SAP: AI Commitments

Both commercial paths into SAP are increasingly bundling AI rather than treating it as an upsell. RISE and GROW customers are now looking at a more concrete expectation to activate Joule Assistants within the first year of their contract, rather than AI being a “get to it eventually” line item. Whichever path you’re on, this changes the planning conversation: AI readiness needs to sit in your migration or adoption plan from day one, not bolted on afterwards. It’s also one more reason a customised, non-standard core makes AI harder and more expensive to get value from (see our guide to SAP clean core).

SAP Business Data Cloud and the Databricks Partnership

SAP continues to widen where Business AI can pull data from. Following last year’s Snowflake integration, SAP Business Data Cloud has extended its ecosystem connections to bring in Databricks, giving customers more flexibility over where their data and AI workloads actually run while keeping SAP’s governance and business context intact. The pattern here is consistent: SAP isn’t trying to be the only place your data lives, it’s trying to be the layer that makes data meaningful wherever it lives.

The 2027 ECC Deadline: Still the Most Urgent Trend

For all the AI headlines, the most commercially urgent trend in the room is one that hasn’t changed: SAP’s mainstream maintenance for ECC ends in 2027, with extended support running to 2030 (and a further, conditional option out to 2033 for some large enterprise customers). Every AI announcement above assumes you’re either on S/4HANA already or heading there. If you’re not, that’s the trend that should be driving your roadmap before any of the others.

What This Means for SAP Customers

Pulled together, the direction is clear: AI is no longer a bolt-on decision, it’s baked into the commercial and technical path SAP is putting everyone on. The practical takeaways:

• If you’re evaluating RISE or GROW, assume Joule activation is part of the deal, not an option to skip
• If your core is heavily customised, that’s now a blocker to both your S/4HANA migration and your ability to get value from BAIP, not two separate problems
• Joule Studio’s move to GA means custom agent development is now something to plan seriously, not just watch
• The 2027 deadline hasn’t moved and shouldn’t be crowded out by the AI headlines