At the center of this work is a research team led by Professors Saji George from the Department of Food Science and Agricultural Chemistry and Suha Jabaji from the Department of Plant Science. The research does not suggest that cannabis products are universally unsafe, nor does it imply that legalization itself has failed. Instead, it highlights a scientific blind spot that may require regulators and producers to rethink how contamination risks are assessed and managed.

The concern revolves around fungi and the toxic compounds some fungi produce, known as mycotoxins. These toxins are already well known in agriculture and food safety, where they can accumulate in crops such as grains, nuts, and spices under favourable environmental conditions. In high enough concentrations, certain mycotoxins have been linked to immunosuppression, neurological effects, and organ damage.

Cannabis plants, particularly their dense flowering buds, can create ideal environments for fungal growth. Warm temperatures, humidity, and post-harvest storage conditions all influence whether microbes establish themselves on the plant. Some fungal species can remain dormant or undetected while still leaving behind toxic compounds that survive even after the fungi themselves are eliminated.

To reduce microbial contamination, cannabis producers commonly rely on gamma irradiation, a decontamination process that subjects products to controlled radiation. The method is widely used in food and medical industries and is considered effective at lowering bacterial and fungal counts without significantly affecting the product itself.

The new study set out to determine how effective gamma irradiation truly is in controlling contamination in cannabis products. Researchers analyzed dried cannabis flower samples using several different scientific approaches. The samples included both irradiated cannabis obtained through a licensed producer collaboration and commercially available cannabis products from multiple producers. Traditional culture-based techniques were used to identify living microbes capable of growing in laboratory conditions. Molecular tools such as PCR and qPCR were used to detect genetic material associated with toxin-producing fungi. The team also employed immunological testing methods to identify the presence of mycotoxins within the samples.

What they discovered was surprising. Although gamma irradiation substantially reduced microbial loads, particularly bacteria, several viable fungal species still persisted after treatment. These included members of Aspergillus, Penicillium, and Fusarium, all generally associated with the production of major toxins. Even more concerning was the fact that residual mycotoxins were still detectable in products compliant with Canadian regulatory standards. These included aflatoxins and ochratoxins, which are currently regulated in cannabis, as well as DON and T-2 toxins, which are not routinely included in Canadian cannabis mycotoxins limit but may still raise health concerns.

In other words, a cannabis product could technically pass current compliance-based testing while still carrying biologically relevant contaminants. The findings challenge a long-standing assumption embedded within current regulations, particularly the heavy reliance on microbial testing methods based on colony-forming units, or CFUs, which measure how many microorganisms can grow under laboratory conditions. But fungi and their toxins do not always behave in straightforward ways. Some microbes may become non-culturable while still leaving behind mycotoxins or other toxic metabolites. Others may survive decontamination processes in dormant forms that evade standard testing.

Mohammad Jamil, one of the researchers involved points out that the gap between compliance and potential biological risk deserves much greater attention and that focusing only on end-stage testing may overlook some contamination pathways and may therefore underestimate overall contamination risk.

Mamta Rani, one of the senior authors, mentioned that contamination cannot always be fully captured through a single testing method. Culture-based screening, molecular diagnostics, and toxin detection techniques each reveal different dimensions of microbial risk. Together, these complementary approaches can provide a more complete picture and support a more comprehensive evaluation of contamination risk.

These hidden risks are especially significant for medical cannabis users. Many patients who rely on cannabis for symptom management may already be immunocompromised because of cancer treatments, organ transplants, autoimmune diseases, or chronic illnesses. For these individuals, even low levels of mycotoxigenic fungi could pose serious risks.

This concern is amplified by how cannabis is consumed. Because a substantial proportion of cannabis is consumed by smoking or vaporizing, fungal spores, toxins, and other contaminants can be delivered directly into sensitive respiratory tissue in the lungs. Past case reports have linked cannabis use to fungal infections such as pulmonary aspergillosis, including in medically vulnerable patients.

Dr George argues that the cannabis industry may need to shift its focus away from relying primarily on final-product testing and post-harvest remediation strategies.

He mentioned that once mycotoxins are formed, they can persist even if microbial counts later decline. Additionally, in some cases, even after degradation, they may still pose a health risk because they can retain some toxicity. Instead, future safety systems may need to emphasize upstream prevention, particularly during cultivation. Certain agricultural sectors already approach mycotoxin control this way, recognizing that preventing fungal establishment early during production is often more effective than attempting to eliminate toxins after they have formed.

The researchers also highlight biological control approaches as an alternative to chemical interventions and post-harvest remediation. The growing acceptance of such strategies reflects a broader movement in agriculture toward more sustainable and resilient production systems. In cannabis, the researchers highlight that beneficial microbial-based formulations are well-suited tools for proactively reducing fungal and mycotoxin-related risks.

Importantly, the researchers are not calling for alarmism. Their work instead advocates for a more nuanced understanding of how safety is defined in a legalized market. Legalization has created systems that provide traceability, standardized oversight, and reduced reliance on unregulated products. Yet the biological complexity of fungi may exceed what current regulatory frameworks were originally built to address. As the cannabis industry matures, standards may need to evolve alongside the science, placing continued responsibility on regulators, producers, and researchers to reassess what safety means in practice.

Legal cannabis systems depend on consumer trust, and that trust can only be maintained when safety standards continue to evolve alongside new scientific evidence, with consumer protection remaining a central priority. That process is common in medicine, food safety, and environmental health. Scientific understanding advances incrementally, often by identifying gaps that were previously invisible. In this case, the research shines a light on an underexamined area of cannabis safety that could shape future policy discussions in Canada and beyond.

The cannabis industry itself may also benefit from such scrutiny. Producers increasingly compete on quality, transparency, and consumer confidence. More advanced contamination prevention systems could become part of that evolution, helping companies differentiate themselves while improving public health protections.

At McGill University, Professor Saji George’s research team is continuing to explore this preventive direction through microbial-based formulations. Over the past several years, the team has identified beneficial bacterial candidates and used them to develop formulations aimed at promoting plant growth, improving crop resilience, and suppressing pathogens. These approaches aim to support crop performance while offering more sustainable and holistic solutions that can help reduce reliance on synthetic fertilizers and pesticides. Professor George has also emphasized the importance of formulation stability, including the use of nanotechnology-based strategies to improve product performance and extend shelf life, a common limitation for scaling microbial products in commercial agriculture.

The team has already applied this work in controlled greenhouse production systems as well as in outdoor field settings with Canadian farmers across different agronomic crops, with promising results in plant growth, crop resilience, and pathogen suppression.

More recently, their attention has turned to cannabis production, where microbial contamination and mycotoxin-related risks create unique challenges for both producers and regulators. The team is evaluating whether similar microbial-based strategies can improve plant health, reduce disease pressure, and help limit contamination risks earlier in cannabis production systems.

Early results have been encouraging, showing improvements in pathogen suppression and overall crop performance. The team is now working to test and optimize these microbial-based formulations in collaboration with licensed producers, with the goal of developing practical tools that can support healthier cannabis crops while reducing contamination risks earlier in the production chain.

As cannabis use continues to expand globally, questions surrounding microbial contamination, toxin persistence, and long-term exposure are unlikely to disappear. If anything, they may become more important as medical cannabis programs grow and vulnerable consumers remain an important user group.

The new findings serve as a reminder that legalization is not the end of scientific inquiry. It is the beginning of a much larger conversation about how society regulates biological products that intersect with medicine, agriculture, commerce, and public health all at once.