What Are GMOs: Genetic Modification, Safety Research, and the Debate

What Genetic Modification Actually Means

A genetically modified organism (GMO) is any organism — plant, animal, bacterium, or fungus — whose genetic material has been altered using genetic engineering techniques that do not occur through natural mating or natural recombination. The definition encompasses a range of techniques, from the insertion of one or a few known genes from another organism into a host genome to the targeted deletion or editing of specific sequences within the host's own DNA. What distinguishes GMOs from conventionally bred crops is not the fact of genetic change — all agriculture involves selection for genetic traits — but the directness, precision, and source of the genetic change.

The first commercial GMO food product was the Flavr Savr tomato, approved in the United States in 1994, which was engineered to slow the softening that occurs after harvest by suppressing a cell wall-degrading enzyme. While the Flavr Savr was commercially unsuccessful, it opened regulatory pathways for subsequent crops. By the late 1990s, herbicide-tolerant soybeans and corn engineered to express insecticidal Bt proteins had been adopted at scale across the United States. Today, around 90% of the corn, soy, cotton, and canola grown in the United States is from genetically engineered varieties.

The techniques of genetic modification have evolved dramatically since the 1990s. First-generation GMOs used Agrobacterium-mediated transformation or biolistics ("gene guns" that fire DNA-coated particles into cells) to insert transgenes at essentially random locations in the host genome. Second-generation techniques used zinc finger nucleases and TALENs — engineered proteins that cut DNA at specific sequences — to direct insertion or deletion to specific locations. The current generation of tools, led by CRISPR-Cas9, enables editing of specific sequences with unprecedented precision, efficiency, and low cost, and is expanding rapidly from research applications into commercial products.

Major Crop Applications and What They Do

Herbicide-tolerant crops, primarily engineered to resist glyphosate (Roundup), were the first widely adopted GM crop trait. By inserting a modified version of the gene encoding EPSPS — the enzyme inhibited by glyphosate — derived from a naturally glyphosate-resistant soil bacterium (Agrobacterium strain CP4), developers created crops that could withstand direct glyphosate application. This allowed farmers to use glyphosate as a broad-spectrum weed control agent in crop fields, simplifying weed management and enabling no-till or reduced-tillage agriculture that preserves soil structure and reduces erosion.

Bt crops express one or more insecticidal proteins originally identified in Bacillus thuringiensis, a soil bacterium used in organic farming as a topical insecticide for decades. When incorporated into the plant genome and expressed in leaf, stem, or grain tissue, Bt proteins protect against specific insect pests. Bt corn controls corn rootworm and European corn borer; Bt cotton controls bollworm. Because the insecticidal protein is expressed within the plant, it reaches stem borers and below-ground pests that topical sprays cannot reach, and reduces the need for broad-spectrum insecticide applications that harm non-target insects.

Biofortified GM crops address nutritional deficiencies in populations dependent on single staple crops. Golden Rice, engineered to produce beta-carotene in the grain endosperm by expressing two genes from the beta-carotene biosynthesis pathway, was developed to address vitamin A deficiency — which causes preventable blindness in millions of children annually in developing countries. After decades of development, regulatory delays, and activist opposition (including the destruction of trial plots in the Philippines), Golden Rice finally received regulatory approval in Bangladesh in 2021. Similarly, GM cassava with enhanced iron and zinc content and virus resistance is under development for sub-Saharan African farmers.

The Safety Evidence

The scientific consensus on the safety of currently approved GM foods is clear and consistent across major scientific organizations: they are as safe to eat as conventional foods. This conclusion is supported by the National Academies of Sciences, Engineering, and Medicine (in a comprehensive 2016 report reviewing over 900 studies), the World Health Organization, the American Medical Association, the European Commission (which funded more than 130 research projects on GM safety over 25 years), and virtually every other major scientific body that has examined the evidence.

The regulatory approval process for GM crops in major markets requires extensive safety testing. In the United States, this involves review by the FDA (food safety), EPA (if the crop produces a pesticidal substance, as Bt crops do), and USDA (environmental release). Applicants must demonstrate that any new protein is non-toxic and non-allergenic, that the nutritional composition of the food is substantially equivalent to conventional counterparts, and that there are no unexpected effects from the transformation process. The EU has an even more stringent regulatory framework managed by the European Food Safety Authority (EFSA), though EU approval of GMOs for cultivation has been extremely rare due to political rather than scientific obstacles.

Claims that GMOs cause cancer, organ damage, or other health effects — most prominently advanced in a 2012 paper by Gilles-Éric Séralini claiming that Roundup-Ready corn caused tumors in rats — have been retracted or thoroughly rebutted. The Séralini paper was retracted by its journal in 2013 following methodological criticism; it used an inappropriate rat strain (Sprague-Dawley rats have a very high spontaneous tumor rate), had grossly insufficient group sizes, and used non-standard statistical analysis. Re-analyses using appropriate methods showed no significant difference in tumor rates. Multiple subsequent studies by independent researchers using standard protocols have not replicated the claimed effects.

Ecological Concerns and Herbicide Resistance

While the human health evidence is settled, the ecological effects of large-scale GMO agriculture raise legitimate concerns that deserve serious analysis. The most significant documented problem is the evolution of herbicide-resistant weeds — "superweeds" — in fields where glyphosate-tolerant crops are grown repeatedly. By applying the same herbicide broadly every season, farmers created intense selective pressure for any weed variants with natural or evolved glyphosate resistance. Over 40 weed species now have glyphosate-resistant populations, and they have spread to hundreds of millions of acres across North America and globally.

The response from the agricultural industry — herbicide-stacking, which introduces resistance to multiple herbicides (including dicamba and 2,4-D) into the same crop — has, predictably, accelerated the evolution of multi-herbicide-resistant weeds. This arms race is a consequence not of genetic modification per se but of the agricultural practice enabled by herbicide-tolerant crops: monoculture with a single weed-management tool. Integrated weed management — combining tillage, crop rotation, cover crops, and herbicide diversity — can slow resistance evolution but requires more agronomic complexity than many large-scale operations are equipped or incentivized to practice.

Bt resistance has also emerged in major pests, including corn rootworm in the United States. Resistance management plans require farmers to plant "refuge" areas of non-Bt crops adjacent to Bt fields, providing a population of susceptible insects that interbreed with any resistant individuals and slow resistance evolution. Compliance with refuge requirements has been inconsistent, and in some regions resistance has advanced more rapidly than models projected. The EPA has progressively tightened refuge requirements and approved stacked multi-trait Bt crops as resistance management tools, but the long-term sustainability of Bt crops as a pest management strategy is genuinely uncertain.

Economic and Power Dynamics

The economics and market structure of GMO seed development are at least as contentious as the ecology. The development of GM crops requires hundreds of millions of dollars in research, regulatory approval, and commercialization costs, which has concentrated the industry in the hands of a small number of large corporations — most prominently Bayer (which acquired Monsanto in 2018), Corteva (the merged DowDuPont agricultural division), and ChemChina-owned Syngenta. These companies hold extensive patent portfolios over both traits and enabling technologies, and license these patents to seed companies under contractual terms that restrict farmers from saving and replanting seed.

The seed-saving restriction — encapsulated in technology use agreements that farmers must sign to purchase patented GM seed — represents a fundamental change in the relationship between farmers and their seeds. For most of agricultural history, farmers saved seed from their harvest to plant the following season, a practice that was both economically rational and a form of agricultural independence. The combination of patent law and contract terms now makes this illegal for patented varieties, creating dependence on annual seed purchases. Critics argue this transfers economic power from farmers to corporations; proponents counter that the investment required to develop improved varieties could not be recouped without intellectual property protection.

For small-scale farmers in developing countries, the picture is complex. Studies of Bt cotton adoption in India showed initial substantial yield increases and reduced pesticide costs, but subsequent pest adaptation and secondary pest emergence have complicated the long-term picture. Many developing-country farmers are now trapped in a cycle of purchasing patented hybrid and GM seeds annually without the technical extension support to manage resistance and maximize the technology's benefits. The promised democratization of agricultural biotechnology — through public-sector GMO development for crops grown by smallholder farmers in Africa and Asia — has proceeded far more slowly than commercial development for high-volume commodity crops in rich countries.

New Breeding Techniques and the CRISPR Question

CRISPR-Cas9 and related gene-editing tools have blurred the regulatory distinctions between GMOs and conventional breeding. Gene editing can precisely delete, modify, or rearrange sequences within the organism's own genome without introducing DNA from another species — a type of change that could in principle occur through conventional mutagenesis breeding (which uses radiation or chemicals to induce random mutations and has been used without special regulation for decades). Whether such edits should be regulated as GMOs is a live question with diverging regulatory answers globally.

The United States has adopted a relatively permissive stance: gene-edited crops that do not contain any DNA from a pathogen or other organism, and that could have been produced by conventional breeding, do not require USDA approval and are not labeled as GMOs. Several gene-edited crops — waxy corn with altered starch composition, high-oleic soybean, mushrooms resistant to browning — have already reached market in the United States under this framework. The EU, by contrast, ruled in 2018 that CRISPR-edited organisms are subject to the full GMO regulatory framework, a decision widely criticized by scientists as scientifically incoherent but reflecting the EU's precautionary regulatory culture.

The distinction matters enormously for agriculture's capacity to respond to climate change. Gene editing offers the potential to rapidly introduce drought tolerance, heat resistance, disease resistance, and improved nutrition into elite crop varieties without the lengthy backcross breeding programs required for conventional trait introgression. A CRISPR-edited version of a popular wheat variety with enhanced drought tolerance could, in principle, be developed in a few years compared to decades of conventional breeding. If these products are regulated as full GMOs — with all the associated costs and political obstacles — the innovation pipeline will continue to be dominated by the few companies that can afford the regulatory burden. If they are treated as equivalent to conventional breeding products, a much wider range of public and private actors can contribute to agricultural innovation.