IN Brief:
- EFSA has organised its future scientific needs around six interconnected regulatory science themes.
- Priorities include novel product assessment, new testing methods, combined exposure, One Health, and emerging risk preparedness.
- Data interoperability and human led use of artificial intelligence are expected to play a larger role in future assessment.
The European Food Safety Authority has set out six interconnected scientific priorities intended to shape how European food risk assessment responds to new products, new technologies, and faster moving evidence over the coming years.
The programme covers innovation and competitiveness; methodological and technical advances; sustainability and One Health; preparedness and rapid response; risk communication and social science; and artificial intelligence and data. EFSA treats the themes as connected rather than ranking them, reflecting the way emerging food risks increasingly cross established scientific and regulatory boundaries.
Novel foods and manufacturing technologies feature prominently. New proteins, biotechnology processes, advanced materials, and other emerging products can generate evidence that does not fit neatly into methods designed around more established food production. Assessors may have to consider unfamiliar biological hazards, processing residues, contaminants, exposure patterns, and production systems while maintaining a consistent standard of evidence.
EFSA also wants greater use of new approach methodologies, omics, exposure science, quantitative adverse outcome pathways, and other techniques capable of strengthening risk assessment or reducing dependence on conventional animal testing. Adoption will depend on validation, however, because a promising analytical method has limited regulatory value until assessors can establish where it works, where it does not, and how its uncertainty should be interpreted.
Combined exposure remains another difficult area. Consumers encounter mixtures of substances rather than one chemical at a time, yet the quantity and quality of evidence can vary sharply between compounds. Developing workable methods for those assessments requires enough sophistication to reflect real exposure without creating models so complex that they cannot be used consistently in regulatory decisions.
The One Health theme places food safety alongside animal health, plant health, environmental conditions, and public health. Zoonotic disease, antimicrobial resistance, contamination events, and climate driven changes rarely respect administrative boundaries, making shared surveillance and interoperable information increasingly important to early warning and response.
Preparedness follows the same logic. A food safety system built mainly around known hazards can be slow when an unfamiliar contaminant, production technology, or biological threat appears. EFSA is seeking stronger foresight, faster mobilisation of expertise, and methods that can absorb new evidence without abandoning the standards of transparency and scientific scrutiny expected in regulatory assessment.
The direction of travel also affects companies developing regulated products. Fermentation derived ingredients, novel proteins, new packaging materials, and other technologies may increasingly be assessed using mechanistic information, alternative testing approaches, larger datasets, and more structured digital evidence. Better methods could make some assessments more efficient, but only where the evidence remains reproducible and its limitations are understood.
Artificial intelligence sits inside that effort as a support tool rather than a substitute for scientific judgement. EFSA envisages increased use of digital systems for evidence retrieval, horizon scanning, predictive modelling, data processing, and communication while maintaining human oversight of the conclusions drawn from those systems.
That distinction becomes critical as the evidence base grows. Faster retrieval is useful only if assessors can trace where information came from, why it was selected, how it was weighted, and whether an automated process introduced systematic bias. A quicker opaque assessment would simply exchange one regulatory bottleneck for a credibility problem.
Data quality and interoperability therefore sit beside artificial intelligence in the programme. EFSA already receives monitoring and analytical data from Member States and food business operators in prescribed formats so that results can be compared and reused. Future digital assessment will depend increasingly on those records being structured, discoverable, and sufficiently consistent to move between systems.
Risk communication forms its own priority because technically sound assessment can still fail if uncertainty is communicated badly or if information arrives after inaccurate narratives have taken hold. EFSA wants social science, audience testing, and behavioural evidence to play a greater role in how risks are explained without allowing popularity or public sentiment to replace scientific assessment.
The priorities will also shape EFSA’s Safe Food, One Future conference in Parma from 29 September to 1 October 2027, when the authority will mark its 25th anniversary and launch Strategy 2034. The conference themes mirror the six areas set out in the new regulatory science work.
The paper does not create an immediate new compliance regime. Its importance lies in showing which capabilities EFSA believes will be needed as food technology, evidence generation, and risk patterns change.
Success will be measured less by the number of new scientific tools named in strategy documents than by whether they become validated methods, interoperable datasets, usable guidance, and faster responses when the next unfamiliar risk or technology reaches the European food system.