A philanthropic organisation focused on sustainable food systems has secured $2 million from the Bezos Earth Fund to develop artificial intelligence tools that enhance the flavour and texture of plant-based proteins.

Food System Innovations (FSI) is amongst fifteen global teams selected for the fund's AI for Climate and Nature Grand Challenge, a $100 million initiative harnessing artificial intelligence to address climate change and biodiversity loss.

The funding will support FSI's NECTAR programme, which has compiled the world's largest publicly accessible database on the sensory characteristics of plant-based meat products and balanced protein foods. The project aims to accelerate consumer adoption of sustainable proteins by improving their taste profiles.

Livestock agriculture accounts for approximately 18% of global greenhouse gas emissions, 27% of freshwater consumption, and 41% of tropical deforestation, according to the organisation. Sustainable proteins represent one of the most effective carbon mitigation strategies in food production—second only to reforestation—and offer the highest carbon dioxide equivalent savings per dollar invested compared to decarbonisation measures across all sectors.

"Taste is the gateway to consumer adoption," said Caroline Cotto, NECTAR's director and co-principal investigator. "NECTAR's data helps the sustainable protein industry refine formulations and enhance flavour, making climate-friendly foods truly irresistible."

The team is developing algorithms that predict sensory attributes and optimise ingredient formulations for sustainable proteins. By combining NECTAR's sensory data with molecular flavour databases, researchers will build an AI model connecting molecular structure, flavour, texture and consumer preference—potentially accelerating product development and market penetration.

Anna Thomas, the project's technical lead and computer science doctoral candidate at Stanford University, noted that early research demonstrates large language models can help revise formulations based on sensory feedback. "With this grant, we can deliver actionable insights that improve taste and speed the protein transition," she stated.

The Bezos Earth Fund launched its AI for Climate and Nature Grand Challenge in 2024, connecting innovators with resources and mentorship from technology partners including Amazon Web Services, Google.org, Nvidia Corporation, Microsoft Research and others.

"These projects show how AI, when developed responsibly and guided by science and local knowledge, can strengthen environmental action and ensure its overall impact on the planet is positive," said Dr Amen Ra Mashariki, director of AI and data strategies at the Bezos Earth Fund.

FSI describes itself as a philanthropic impact platform investing in humane and sustainable food futures. The organisation functions as both funder and operator, directly implementing programmes whilst supporting others through grants, investments and capacity building.

The initiative addresses a persistent challenge facing plant-based protein manufacturers: whilst environmental and health arguments favour sustainable alternatives, widespread consumer adoption ultimately depends on matching or exceeding the taste and texture of conventional meat products.

Despite substantial investment in plant-based protein companies over recent years, market growth has disappointed some expectations as early adopters proved insufficient to drive mass-market penetration. Industry observers increasingly recognise that taste parity—or superiority—represents a prerequisite for broader consumer acceptance rather than a secondary consideration.

AI-assisted formulation development could potentially accelerate the iterative process of improving plant-based products, which traditionally relies on extensive consumer testing and gradual refinement. Whether machine learning models can effectively predict human sensory preferences remains to be demonstrated at scale.

The project reflects growing interest in applying artificial intelligence to food science challenges, from agricultural optimisation to product development. Success could establish methodologies applicable beyond plant-based proteins to other sustainable food innovations.