Under the EU’s new rules on New Genomic Techniques (NGTs), most New GMOs will reach the market without being labelled. For Non-GMO and organic producers, whose standards explicitly exclude them, that raises a practical question: how can you make sure what’s in your supply chain?
That is the question the DARWIN project is working on. ENGA Board Chair Florian Faber, who is also Director of ENGA member ARGE Gentechnik-frei, attended the project’s policy workshop and spoke to project coordinator Odd-Gunnar Wikmark and Dr Nancy Roosens, Head of Division Biotechnology at Sciensano and a leading researcher with DARWIN.
What DARWIN is
DARWIN is an EU-funded research project coordinated by the Norwegian research institute NORCE. It is developing methods to detect, identify and trace plants and products made with NGTs, and digital tools to support traceability. Its partners include research institutes, laboratories, IFOAM Organics Europe and VLOG. The project focuses on two crops, rice and tomato. That limits what its methods can be used for today. Wikmark is clear about this, describing DARWIN as “the beginning of showing the path”.
Why New GMOs are harder to find
Testing for conventional GMOs relies on inserted DNA. Laboratories screen for common genetic elements, then use the unique junction where the new DNA sits in the genome to identify the specific GMO.
NGT plants can look very different. An edit can be a single DNA base, with no foreign DNA left behind. The same change might also occur naturally or through conventional breeding. Roosens sums up the problem: detecting an edit does not prove where a plant came from. A laboratory may find the change, but it still needs more information to say whether it came from an NGT plant.
What the research has shown
Most of the results so far come from rice and are proofs of concept:
- Very small edits can be found. Researchers detected a single-letter change in genome-edited rice, even in very small amounts of DNA and a first trial across several laboratories has been completed.
- Edited plants can be told apart. Looking at the edit together with other markers in the plant’s DNA creates a kind of genetic fingerprint. This helps distinguish an edited rice line from other plants with the same change.
- Mixtures can be tested. NGT rice was identified in a sample that was 99% soybean and only 1% NGT rice.
The research also demonstrates the growing role of artificial intelligence. AI-based tools can help researchers analyse large volumes of genomic information, identify relevant genetic markers and accelerate the development of detection strategies. Roosens explained that DARWIN uses existing AI tools rather than developing its own and that Sciensano plans to develop a further method next year, based on laser capture microdissection and whole genome sequencing.
The implications are significant: analytical approaches that previously required extensive manual examination of genomic data can increasingly be supported by automated computational analysis.
Can it be affordable?
Roosens is cautiously optimistic. Sequencing a plant genome costs around €1,000, she said. Once a laboratory has that reference data, it can develop a simple, low-cost PCR test for routine use. Enforcement laboratories could then save whole genome sequencing for cases that need confirmation.
Digital traceability
DARWIN is also building a digital platform that uses a permissioned blockchain to record each step from seed to shelf. Lab results can be linked to the batches they relate to. The platform is at an early stage and the workshop demonstration used illustrative data.
What the workshop debated
The workshop also included a panel with leading representatives of EU institutions, laboratories, organic and Non-GMO certification, farming and the food industry. The discussion took place under Chatham House Rules and therefore what was said is not be attributed. Several themes stood out:
- Testing is one tool among several. Detection can show that a change is present, but not how a plant was produced. Speakers pointed to combining analysis with documentation and digital records, where a gap in the record can trigger a laboratory test.
- Standards are still to be set. Current performance requirements for detection methods cover PCR only, so requirements for sequencing-based methods may be needed in the coming months. There was also discussion of whether some requirements may need to be less strict for plants with very small changes, with other tools filling the gap.
- Detection for NGT1 is not mandatory. Developers of category 1 plants will not have to supply a detection method, although seeds will be labelled. Whilst analytical tools offer reliable, trustworthy possibilities for standard verification, both for the organic and Non-GMO sectors, there was discussion around the need for more information to develop methods. This will enhance the credibility of non-GMO and organic certification systems.
- Who pays? Some participants said the burden so far falls on those who do not want to use NGTs. Non-GMO certification will also need to start earlier in the supply chain, adding complexity and cost. To achieve a more balanced distribution of burdens, it should be possible to regulate NGT1 cultivation through national coexistence rules that do not conflict with the NGT Regulation.
- Patents. Participants linked detection to patent concerns, because it will be hard to tell NGT1 plants from conventionally bred ones.
What still needs to happen
The DARWIN team highlighted the gaps for NGT detection.
Reference data: Every speaker said laboratories need access to whole genome sequence data from NGT plants. This does not mean public disclosure. A database with controlled access could give competent authorities and designated laboratories the information while protecting commercial confidentiality. Documentary traceability remains essential alongside testing.
More crops: Methods built for rice and tomato will not automatically work for other crops. Rice and tomato make up about 45% of entries in a September 2025 snapshot of the EU-SAGE database of genome-edited crops, so more than half the entries involve other species.
Validation: The methods are not yet ready for routine enforcement. The next step is validation across laboratories under ENGL (European Network of GMO Laboratories) performance criteria.
Incentives and funding: Wikmark noted that NGT1 plants will not be regulated like GMOs. So, he said, “unless there is an incentive” to extend the work, it may not happen. A way forward here may be a project with the organic sector to target the crops it cares about most. Sciensano also warned that verification capacity could decline without sustained investment in laboratories, equipment and training.
What it means for the Non-GMO sector
DARWIN shows that independent testing for New GMOs is technically possible. But as the project’s own researchers stress, it will only work in practice with access to reference data, validated methods, funded laboratories and good documentary traceability.
The European Commission is now preparing the detailed rules that will set the information requirements for NGT plants. Those rules will largely decide how much of this science can be used in practice.
Further information: DARWIN project | DARWIN policy workshop, 22 October 2026