Gene-edited plants and the EU: what Parliament voted for, and what happens next
· 7 min read · by Henrik Aalto
On 7 February 2024 the European Parliament voted on how the EU should regulate plants made with gene editing. The result was close: 307 in favour, 263 against, 41 abstentions. It was also not a law. What MEPs adopted was a set of amendments to a Commission proposal, and the file went straight back to committee so that negotiations with the member states can start, once the member states have a position of their own. They do not have one yet.
That has left plenty of people unsure what has actually been decided. Here is what is on the table, what the Parliament changed, and what still has to happen before anything applies on a farm.
How Europe got here
The EU's GMO rules date from 2001. They regulate organisms according to the process used to make them, and they exempt plants from conventional mutagenesis, the decades-old practice of exposing seeds to radiation or chemicals and selecting useful mutants, on the grounds of a long record of safe use.
When CRISPR arrived, breeders hoped targeted mutagenesis would fall under the same exemption. In July 2018 the Court of Justice of the EU said no. Organisms obtained with new mutagenesis techniques are GMOs and must go through the full authorisation procedure. Since that procedure takes years and costs millions, almost nobody has tried it with an edited crop.
The Commission then ran a study, published in April 2021, which concluded that the 2001 framework was not fit for purpose for these techniques. Its legislative proposal followed on 5 July 2023 (COM(2023) 411).
What counts as a "new genomic technique"
The proposal covers two families of methods.
Targeted mutagenesis. A guided enzyme such as Cas9 cuts the DNA at a chosen site. The cell repairs the break, often imperfectly, and the small deletion or insertion left behind can switch a gene off. If a short repair template is supplied, the cell can copy a specific small change instead. No foreign DNA need remain in the final plant.
Cisgenesis and intragenesis. A gene, or rearranged pieces of genes, from the same species or a crossable relative is inserted. A potato might receive a blight-resistance gene from a wild potato, something a breeder could in theory have crossed in over many generations of back-crossing.
Transgenesis, moving genes between species that cannot interbreed, is not covered. Those plants stay under the existing GMO rules.
Two categories
The heart of the proposal is a split.
Category 1 NGT plants are those considered equivalent to conventional plants. The test sits in an annex: a plant qualifies if it differs from its parent by no more than 20 modifications of listed types, such as substituting or inserting up to 20 nucleotides, deleting any number of nucleotides, or inserting cisgenic material. The reasoning is that changes of this size and kind could also arise through natural mutation or ordinary crossing.
Category 1 plants would go through a verification procedure, not a risk assessment. A national authority checks the criteria, other member states and the Commission can object, and if all is in order the plant goes into a public database and is treated like any conventional variety.
Category 2 covers every other NGT plant: more extensive edits, more changes than the threshold allows, larger cisgenic constructs. These would stay under GMO legislation, but with risk assessment, detection and monitoring requirements adapted to the plant's profile, and with some regulatory incentives for traits that serve sustainability goals.
Whether the line of 20 modifications has a sound scientific basis has been debated since the proposal appeared. It is, at heart, a policy number dressed in molecular language. That does not make it wrong, but it does mean it will be argued over.
The comparison
The table follows the Commission's proposal. Where the Parliament has asked for something different, that is noted.
| Category 1 NGT | Category 2 NGT | Transgenic GMO | |
|---|---|---|---|
| Before market | Verification of criteria by a national authority | Authorisation under GMO law | Authorisation under GMO law |
| Risk assessment | None | Yes, adapted to the plant's risk profile | Full assessment involving EFSA |
| Detection method | Not required | Required, adapted where testing is not feasible | Event-specific method required |
| Food and feed label | None (Parliament: "new genomic techniques" label) | GMO label, optionally naming the trait | GMO label; accidental traces under 0.9% exempt |
| Seed label | Yes, as category 1 | Yes | Yes |
| Organic farming | Excluded | Excluded | Excluded |
| National cultivation bans | No | Allowed | Allowed |
| Public register | Yes, online database | Yes | Yes |
What the Parliament changed
MEPs accepted the two-category structure but added conditions that make category 1 a good deal heavier than the Commission intended.
- Labelling and traceability. Category 1 plants and products should carry a "new genomic techniques" label, and the information that a product contains NGT plants should travel with it along the supply chain.
- Organic. Both categories stay out of organic production. Accidental, technically unavoidable presence of category 1 material would not count as a breach of organic rules.
- Herbicide tolerance. Plants made tolerant to herbicides would be excluded from category 1.
- The threshold. The Parliament amended the rules on the size and number of modifications allowed for category 1.
- Monitoring. Category 2 plants would always need an environmental monitoring plan when first authorised.
- Patents. The headline amendment. NGT plants, plant material, parts of them, the genetic information they contain and the process features they carry should not be patentable. Applicants for category 1 status would also have to disclose any patents or pending applications covering the plant. The Commission is asked to report by June 2025 on how patents affect breeders' and farmers' access to plant material, with a legislative proposal to follow.
The patent problem
The patent ban is the most popular amendment and the hardest to make work.
Patents on plants in Europe are mostly granted by the European Patent Office under the European Patent Convention, a treaty whose members include states outside the EU. A regulation about food and feed cannot simply rewrite it. EU law does shape how the EPO reads its own rules: the EU's 1998 directive on biotech patents is built into the convention's implementing regulations, and the office has followed EU political signals before, notably over plants produced by conventional breeding. A ban that actually binds would most likely require changes to that directive, and then to the EPO's rules.
The tension underneath is simple to state. The regulation would say a category 1 plant is equivalent to something conventional breeding could produce. Patent law rewards the same plant as a new, inventive, technical result. Breeders, small ones especially, worry about having to search patent databases before they can use a variety in their own crossing programmes. Seed markets are already concentrated, as our look at concentration ratios shows, and broad trait patents could tighten that further.
Europe has been here before with very broad claims on crop plants, a history covered in our piece on the EPO's soybean patent decision. And the plant variety rights system explained in our UPOV 91 explainer was designed to leave breeders free to use protected varieties for further breeding. Patents carry no general exemption of that kind, which is the crux of the breeders' complaint.
What still has to happen
- The Council needs a position. Member states must agree by qualified majority. The Spanish presidency tried and failed in December 2023. The Belgian presidency, in charge since January 2024, has been putting new compromise texts to EU ambassadors. Patents are the main obstacle, and labelling, national opt-outs and the category 1 criteria are not settled either.
- Trilogues. Once the Council has a mandate, Parliament, Council and Commission negotiate a single text behind closed doors. That is where the patent ban, the label and the threshold will be traded against each other.
- Final adoption. Both institutions vote on the agreed text. The regulation then applies after a transition period.
The calendar is unkind. European elections are due on 6–9 June 2024, and a new Commission will follow. The Parliament can protect its work by formally closing its first reading before the elections, so the next Parliament does not have to start again. The Council has no such deadline, and no obligation to agree at all.
Who wants what
Plant breeders' associations and many plant scientists want category 1 kept as close to the Commission's version as possible, arguing that labels and traceability for products indistinguishable from conventional ones would recreate the GMO stalemate under a new name. Organic farmers, many retailers and environmental groups argue the opposite: that without labels and traceability, consumers lose the choice they have had since 2001, and organic farmers lose the ability to keep edited material out of their supply chains. Some governments care most about patents, others about national opt-outs. Positions do not line up neatly along the usual pro- and anti-GM divide, which is part of why the Council is stuck.
Where that leaves us
As of late February 2024 the Commission has proposed, the Parliament has amended, and the Council is still counting votes. Meanwhile the legal position has not moved an inch. Gene-edited plants remain GMOs under EU law, exactly as the court ruled in 2018, and any company wanting to sell one in Europe still faces the full authorisation process.
The vote in Strasbourg was a real step. It was not the finish line, and the hardest part of the argument, over who may own an edited plant, has barely started.
Written by Henrik Aalto
Henrik writes about genetic engineering, gene editing and synthetic biology – the science, the patents and the gap between a press release and a product. He reads the supplementary data so you do not have to.
