By Athira Sreenivasan, Junior LCA Specialist, CarbonZero AB
From my perspective working with LCAs and EPDs, I think the misunderstanding starts with how we frame the role of AI.
It is not a new methodology. It does not change EN 15804. It does not reinterpret ISO 14040–44. It does not decide on system boundaries, allocation rules, or impact categories. The standards remain exactly where they have always been — as the foundation of credible and comparable assessments.
Methodological judgment is still human work. What AI influences is something else: how we handle information.
AI does not change the standards
One of the most common concerns I hear is that AI will somehow dilute rigor. But the structure of an LCA is defined by established standards. Functional units, system boundaries, impact categories, documentation requirements — these do not shift because a new tool enters the workflow.
If anything, the presence of AI makes clarity around methodology even more important. The stronger the framework, the more effectively technology can support it.
AI operates within the rules. It does not redefine them.
What AI actually changes
Where AI does begin to influence practice is in how we handle information.
In reality, much of LCA work is not about debating allocation principles. It is about managing inputs. Supplier documentation arrives in different formats. Data must be structured and validated. Materials, components, transport, and packaging need to be modelled consistently. When one parameter changes, multiple outputs must be updated.
This is where AI becomes relevant.When data extraction becomes faster, when inconsistencies are flagged automatically, and when structured datasets can be reused across products, the workflow shifts. The science remains intact, but iteration becomes easier. Scenario modelling becomes more dynamic. Portfolio-level comparisons become feasible without rebuilding everything from scratch.
That shift does not remove expertise. It reshapes how and where it is applied.
Transparency remains non-negotiable
There is also a legitimate concern around transparency. LCA professionals rely on traceability. Every figure must be explainable. Every emission factor must have a documented source. If AI introduces opacity, it undermines credibility.
But used within a structured system, AI does not have to be a black box. It can support traceability by organizing inputs clearly, identifying gaps early, and ensuring consistency across assessments. Validation, interpretation, and accountability still belong to the practitioner.
AI can assist in structuring and analysing data. It does not replace professional responsibility.
From operational effort to strategic leverage
What I find most interesting is how AI changes the distribution of effort. When repetitive structuring and updating tasks are reduced, more time can be spent on interpretation, improvement strategies, and meaningful dialogue with product and procurement teams. Instead of rebuilding similar models, practitioners can focus on understanding reduction pathways and long-term impact.
That is not a loss of relevance. It is an expansion of influence.
The demand for product-level sustainability data is increasing. Internal teams expect quicker answers and more dynamic insights. In that context, purely manual workflows become difficult to scale. AI does not replace professional judgment, but it can make that judgment more scalable and responsive.
For me, the key is intention. AI should be integrated as a support layer within a standards-based, transparent system. The expertise of the LCA practitioner remains central — not because technology cannot calculate, but because interpretation, context, and accountability cannot be automated away.
The future of LCA practice will likely combine deep methodological knowledge with intelligent data handling. The real question may not be whether AI belongs in LCA — but how we choose to use it.
So as AI becomes more embedded in sustainability workflows, how do you see it shaping your own role as an LCA practitioner?
.jpg)
.png)