{
  "id": 16867704,
  "title": "Why better data could be the key to AI-driven food R&D",
  "url": "https://urgent.news/2026/10/28/why-better-data-could-be-the-key-to-ai-driven-food-r-d",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-28T07:49:21.000Z",
  "source": {
    "name": "YourStory",
    "slug": "yourstory",
    "url": "https://yourstory.com/2026/10/how-genai-digital-ontologies-are-redefining-food-rd-food-safety"
  },
  "original_language": "en",
  "account": "In the realm of food, beverage, and wellness research and development, an increasing amount of data is required to create products that are not only delicious but also meet the evolving safety demands. However, this surge in data has led to challenges in managing and connecting it all effectively. The World Health Organization estimates that in 2021, unsafe food caused 866 million illnesses and 1.5 million deaths globally, highlighting the urgency of addressing this issue.\n\nA potential solution lies in the adoption of shared vocabularies, specifically through the use of ontologies. These structured vocabularies provide a common language for food data, linking different names, specifications, and regulations together. For example, an ontology like FoodOn can connect \"haldi,\" \"turmeric,\" and \"Curcuma longa\" to their respective specifications, supplier lot numbers, test results, claims, and market requirements. This enables AI systems to recognize that multiple names refer to the same ingredient and connect the relevant details.\n\nGenerative AI can assist in various aspects of R&D, such as suggesting alternative ingredients and drafting reports. However, when information is fragmented and unstructured, AI may miss important connections or produce unverifiable answers. By grounding AI in structured, connected facts provided by an ontology, the technology can function more accurately and reliably.\n\nSafety and regulatory compliance can be integrated into the development process through connected data systems. For instance, if a developer chooses a rice-based protein, the system can flag potential inorganic arsenic risks before a pilot batch is made. This proactive approach reduces the likelihood of rework, delays, and costly product recalls.\n\nWhile AI can generate formulations with confidence, it is essential to remember that it cannot taste or feel the nuances of a product. The working rule should be that AI proposes ideas, but scientists must ultimately decide, test, and sign off on them. This ensures accountability for safety and regulatory decisions.\n\nLooking ahead, the future of food R&D does not involve robot chefs but rather shorter development cycles, leveraging AI to streamline information searching and knowledge retention. By investing in the cleaning, naming, and connecting of data, organizations can unlock the full potential of AI in driving innovation and ensuring the safety of their products.",
  "summary": "As food companies navigate increasingly complex product development and safety requirements, connected data and generative AI could help scientists make better-informed decisions, reduce manual work and identify potential risks earlier.",
  "key_points": [
    "Unsafe food caused 866 million illnesses and 1.5 million deaths globally in 2021.",
    "Ontologies like FoodOn provide shared vocabularies for food data.",
    "Generative AI can assist R&D but requires structured data for accuracy."
  ],
  "editors_take": "Standardizing fragmented food data through structured ontologies transforms generative AI from an unreliable assistant into a precise tool for accelerating innovation and preventing costly safety recalls.",
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}