In the news...

In the News: Can AI Make Our Food Supply Safer?

Source: Food & Wine, August 15, 2025

Researchers at the University of South Australia have unveiled a breakthrough that could transform

global food safety. By combining hyperspectral imaging (HSI) with machine learning, their new AI system can detect dangerous mycotoxins in cereal grains and nuts before they reach consumers.

These toxic compounds, produced by fungi, are linked to foodborne illness, cancer, and crop losses. According to the World Health Organization, more than 600 million people fall ill each year from food poisoning, with 4.2 million deaths worldwide. The Food and Agriculture Organization (FAO) estimates that up to a quarter of the world’s crops may be contaminated.

How It Works

• Hyperspectral imaging captures detailed spectral “footprints” of food samples.
• Machine learning algorithms analyze those patterns to detect specific toxins.
• The system can classify contaminated grains and nuts with 90–95% accuracy in lab tests.
• Particularly effective against aflatoxin B1, one of the most carcinogenic foodborne toxins.

Why It Matters

This AI-driven process is fast, non-invasive, and scalable. It could be deployed on industrial processing lines or even handheld devices, helping farmers, food processors, and regulators identify contamination in real time. That means fewer food recalls, safer meals, and reduced trade losses worldwide.

Professor Sang-Heon Lee, project lead at UniSA, noted that the approach is not only accurate but also practical: “It offers a scalable solution for industrial food safety, from sorting almonds to inspecting wheat and maize shipments.”

Looking Ahead

While current success rates are highest in laboratory conditions, researchers are refining the technology for broader real-world use. If adopted widely, this AI could drastically cut down on the illnesses and deaths caused by contaminated foods each year—making our food supply safer than ever before.