Across the industrial programmes we support, the conversations we have with engineering leaders, and the industry gatherings we attend, a clear picture is emerging for 2026: PLM is evolving from its traditional engineering roots into a key pillar of enterprise transformation.
Here is what that looks like in practice.
AI stops being a demo and becomes part of the job
For the past two years, most AI in PLM meant a chat interface bolted onto an existing platform. Impressive in demos, underwhelming in practice. That is changing.
What we are seeing now is AI embedded directly into engineering workflows: not as a layer on top but woven into the daily tasks that used to consume hours of low-value effort. Classifying parts. Flagging reuse opportunities. Drafting change summaries. Detecting duplicate components before they make it into a BOM. These are not glamorous use cases, but they are the ones that move throughput metrics and reduce downstream errors.
The next step is already visible in the more advanced projects: multi-agent systems where different AI agents handle distinct tasks in parallel: change impact analysis, requirements traceability, supplier data enrichment and hand off to each other without waiting for a human to orchestrate. It is less about having a smarter search bar and more about having a team of specialised assistants that never sleep and never drop the ball on routine checks.
The real measure of success is not how impressive the AI looks. It is how many engineering hours get freed from routine work.
Cloud adoption is real, but hybrid remains the dominant model
Cloud PLM adoption is real. Faster deployments, better cross-site collaboration, and lower infrastructure costs are major drivers. But most large enterprises are not starting from a clean slate. They carry legacy CAD and PDM environments, data residency constraints, and supplier ecosystems that span dozens of organisations with their own tools. For them, cloud migration is not a switch, it is a multi-year journey.
The organisations making the most progress are modernising incrementally, keeping what works and fixing what does not. The question has moved on from "cloud or on-premise" to something more pragmatic: how do we govern data consistently across a hybrid landscape?
Composable architectures replace the monolith vision
For a long time, the PLM market sold a simple idea: one platform to rule everything. Design, BOM, change management, document control, quality: one system, one truth.
That idea is fading. Not because the platforms failed, but because the complexity of modern product development has outgrown any single tool's ability to handle it cleanly. What is replacing it is a composable approach: a solid PLM backbone for core data and governance, combined with specialised applications connected through open APIs. Requirements management, systems engineering, manufacturing execution, sustainability compliance: each handled by the best tool for the job, all speaking a common language.
The hard part is not the architecture. It is the governance. Composability only works if you have answered the difficult questions: which system owns which truth? How do changes propagate? Who resolves conflicts? This is where enterprise architecture expertise makes the difference and where we invest a significant part of our work with clients.
The digital thread becomes a buying requirement
"Digital thread" has been a slogan for years. In 2026, it is becoming a buying requirement.
Industrial customers are starting to ask concrete questions: can your PLM reliably connect as-designed to as-planned, as-built, and as-maintained? Not just store the data, but use it to close the loop, so that what comes back from the field actually feeds into the next design decision?
This shift toward semantic digital threads, ontologies, and traceable data lineage is not driven by technology enthusiasm. It is driven by hard operational problems: late-stage failures, recall costs, compliance mandates, and the growing pressure to demonstrate carbon and material traceability across the supply chain. The digital thread is becoming the infrastructure that makes all of that possible.
Sustainability moves from reporting to design constraints
In 2026, sustainability finally appears where it belongs: inside product definition and change processes.
Instead of being treated as an end-of-cycle reporting exercise, sustainability constraints are being embedded directly into specifications, materials choices, packaging rules and regulatory checks.
Learn more about Ecodesign
Supply chain regulations, carbon reporting and material restrictions are turning sustainability into a business-critical risk topic that PLM systems must enforce at the point of decision.
PLM as a strategic business enabler
All of the above is real. The AI, the architecture, the digital thread, the sustainability integration, the technology exists and is being deployed. But the industrial programmes that are actually delivering business outcomes share something that has nothing to do with the tool selection.
They have executive sponsorship. They treat data architecture as a strategic asset rather than an IT project. They invest in change management alongside the implementation. And they measure PLM performance in business outcomes as time to market, cost of quality, throughput, not in modules deployed or licences activated.
Model-Based Systems Engineering, Model-Based Definition, and the broader push toward model-based enterprise are gaining real traction. The new SysML v2 standard is a meaningful step forward for systems complexity.
But the human dimension matters equally: people must be front and centre. Technology augments, it does not replace the engineering judgment and change leadership that make PLM programmes succeed.
Looking ahead
The trends above are reinforcing each other. AI delivers more value when it operates on a well-governed data foundation. The digital thread becomes meaningful when it connects real decisions across the lifecycle. Composable architecture enables both. And none of it lands without the organisational commitment to make it stick.
2026 is a good year to take stock of where your PLM programme stands and where you want it to go.