Can AI communicate with plants in any way shape or form ?

Technically humans can already "communicate" with plants. Give water, plant thrives and signals with flowers. Plant droops, signals to outside world "I need water". Can AI enhance or elevate communication ( not action but interaction ) with plants in a meaningful way ?

Background

Some research groups have explored using machine learning on plant physiological signals to infer abiotic stress, such as drought, salinity or insufficient light, by training models on time-series data from sensors embedded in leaves or soil. Laboratory work has shown that classifiers can distinguish between healthy and stressed states with accuracies above 90% when provided with multispectral reflectance or volatile organic compound measurements, indicating that plants already emit signals that can be translated into meaningful labels. A small number of experiments have coupled these predictions with closed-loop systems that deliver water or nutrients proportionally to inferred need, demonstrating a rudimentary form of interaction rather than one-way communication. Broader efforts in plant phenotyping leverage computer vision and LiDAR to quantify growth rates and morphological changes over time, generating data streams that some researchers frame as a shared language between plant and algorithm. At the same time, other teams have tested bioacoustic approaches, recording ultrasonic emissions that plants release under stress and then using neural networks to convert these into human-readable alerts. These strands collectively suggest that measurable plant responses can be operationally interpreted, but they have not yet established a two-way, semantically rich dialogue.

— Enriched May 15, 2026