Edited by: Stephen Allen Morse, Centers for Disease Control and Prevention (CDC), United States
Reviewed by: Gerald Epstein, National Defense University, United States; Fernanda Rei Leal, University of Trás-os-Montes and Alto Douro, Portugal
This article was submitted to Biosafety and Biosecurity, a section of the journal Frontiers in Bioengineering and Biotechnology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
This article discusses a previously unrecognized avenue for bioterrorism and biocrime. It is suggested that new gene editing technologies may have the potential to create plants that are genetically modified in harmful ways, either in terms of their effect on the plant itself or in terms of harming those who would consume foods produced by that plant. While several risk scenarios involving GMOs—such as antibiotic resistant pathogens, synthetic biology, or mixing of non-GMO seeds with GMO seeds—have previously have been recognized, the new vulnerability is rooted in a different paradigm—that of clandestinely manipulating GMOs to create damage. The ability to actively inflict diseases on plants would pose serious health hazards to both humans and animals, have detrimental consequences to the economy, and directly threaten the food supply. As this is the first study of this kind, the full scope and impact of suck attacks—especially those involving the intended misuse of technologies such as gene-drives—merits further investigation. Herein, the plausibility of some of the new risks will be analyzed by, (1) Highlighting ownership and origination issues (esp. of event-specific GM-plants) as unrecognized risk factors; (2) Investigating the unique role of GMOs, why—and how—certified GMOs could become a new venue for such attacks; (3) Analyzing possible dual-use potentials of modern technologies and research oriented toward the advancement of GMOs, plant breeding and crop improvement. The identification and analysis of harmful genetic manipulations to utilize (covertly modified) plants (GMOs and non-GMOs) as an attack vector show that these concerns need to be taken seriously, raising the prospect not only of direct harm, but of the more likely effects in generating public concern, reputational harm of agricultural biotechnology companies, law-suits, and increased import bans of certain plants or their derived products.
According to their current definition, biological weapons achieve their intended target effects through the infectivity of disease-causing infectious agents. The CDC
While there has been much focus on traditional infectious agents such as bacteria and viruses, this article describes a previously underappreciated vulnerability. It is suggested that novel gene editors and other scientific advances allow for a clandestine manipulation of GM plants
Whether it is consternation about weediness, the development of super-bugs and antibiotic resistance, influences on human health, or concerns about unavoidable and irreversible impacts on “neighboring farmers, regions, and countries” (Brown,
In the context of emerging agricultural technologies via infectious genetically modified viruses engineered to edit crop chromosomes directly in the field, Reeves et al. (
The most skillful attacks don't rely on natural circumstances and happenstance. As detailed by Berns et al. (
This article describes a previously unrecognized form of biocrime—the weaponization of GMOs. This may be best explained by the following—already familiar—predicament of seed contamination which once again made the news. On February 7, 2019, an unauthorized GMO strain was identified in Europe, mixed in with the natural seed bought from Bayer-Monsanto. By the time an official recall was issued, some of the seeds had already been planted, covering 8, 000 ha in France and 3, 000 ha in Germany.
“Unauthorized GMOs” (UGMOs) are GMOs that are released in the market of a certain jurisdiction without prior authorization (Arulandhu et al.,
Instead of accidental and unintended mingling of merely unauthorized GMOs, it is suggested that existing GMOs may be exchanged in clandestine, with the intent to create damage. Thus, instead of attacks on plants, the possibility is considered that malignant genetic manipulations may turn plants themselves into harmful attack vectors. This possibility opens up an unrecognized avenue for bioterrorism or biocrime—either by maliciously modifying a natural plant or (perhaps more perniciously) sabotaging a previously approved GMO. Compounding the problem is that such an introduction will be difficult to detect. This is particularly the case when nobody is looking for such manipulations. However, an additional factor is just as important. It will be shown that current GMO identification and authentication techniques suffer from critical vulnerabilities that may be misused. Additionally, several attack scenarios are analyzed, how perpetrators may attempt to introduce harm, either on the plant itself or in terms of those who would consume foods produced by the plant. Some of these need to be taken seriously, also because of their impact on public perception and acceptability of GMOs and agricultural biotechnology companies.
Biocrime in form of clandestine manipulations of GMOs: A summary of the main contributing factors. Details and feasibility of these vulnerabilities are analyzed in the following sections.
Although GMOs continue to be rigorously investigated and are regulated by The International Regulations and the Codex guidelines (see e.g., FAO/WHO,
Even in areas with rather strict regulations—as in Germany and France—it remains unclear how unauthorized GMOs can enter the food and feed chain. That is, how this commingling of seed happens, accidentally and unintentionally. Although the exact mechanisms are unknown, it may be possible that specific situations could be exploited for nefarious activities, such as differences and gaps in jurisdiction, limits of detection methods, or the deliberate release from field-trials.
An additional concern arises due to limited post-market analysis practices and regulations. While GMOs undergo critical risk assessment before approval, the presence of additional, deleted, or manipulated genes might not be obvious once such approval has been obtained. An attacker may be masquerading a manipulated (and hence, hazardous) as a certified GMO, or threaten to do so.
Traceability and labeling are critical factors to help authenticate GMOs. Currently, this is realized via unique identifiers. According to Bar-Yam et al. (
Although such a code enables a genuine description of the product, this does not automatically validate the authenticity of the GMO itself. Already 15 years ago, it was realized that it would be crucial to have an identifier directly at the genomic level. This was realized via unique flanking sequences (Levine,
The very presence of unique identifiers of a GM plant (via, e.g., the event-specific characterization) might seem to validate the product under investigation. Unfortunately, this is based on a mistaken understanding regarding the functionality of such GMO signatures.
Traditionally, a signature string provides assurance of origination and content of a specific document, just as electronic signature and authentication methods guarantee integrity of an electronic sender and their products. However, with GMOs there is a real problem. Any self-authenticating signature element (even if it is enhanced by cryptographic methods, Mueller,
The problem is not only the manipulation of GMOs. What makes it worse is that practically these can rather easily be hidden inside the genome and that attacks may be done in a variety of settings (see below). Not only could plants be targeted via vectors directly in the field, it would similarly be possible to exploit gaps in the supply chain (Frazar et al.,
Although numerous technologies have been developed to detect foreign DNA in GMO food and feed (see Bai et al.,
To alleviate regulatory concerns, CRISPR-based technology is gradually avoiding using transgene DNA. This is further complicating problems. An attacker could directly manipulate gene expression within a plant through dsRNA based post-translational gene-silencing methods (see the Discussion Section for technical and scientific challenges). Such modifications would be much more difficult to detect. Unless the dsRNA made by the GM plant is intended to act as a pesticide, the RNA itself is rarely formally considered in a risk assessment. As we are still lacking sufficient knowledge about the many novel RNA molecules (Heinemann et al.,
It may be worthwhile to summarize the above (see
Agents seeking to perform the attack learn the event-specific characterization (typically, the flanking/border regions of the transgenes) of a specific GMO. (This information is publicly available).
The attackers wait until critical risk assessments and authentication processes—if existent in that jurisdiction—are completed (or performs the attacks in situations where these are not adequately supported). Alternatively, the attackers mimic some of the ways how UGMOs enter the market (Rostoks et al.,
By utilizing novel technologies (e.g., CRISPR/Cas — see section 4 for scientific, technical, and operational challenges), attackers may exploit unrecognized insights from GMO and agricultural research to introduce various malignancies at the genomic or proteomic level (see
Depending on the intended scope and impact of these manipulations, the attackers can tailor how—and when—they can make their intrusions public (e.g., blackmailing, real threat, or mere hoax, see section 4.5 below and
It is critical to observe that the manipulated GMOs would still carry the authentic identifiers of the originals. Thus, with respect to these official identifiers—as could be verified by independent labs—the manipulated product would pass for the original.
Generic overview of attacks via the clandestine exchange of GM plants with the intent to cause harm. The consequences on the right are roughly ordered from top to bottom in terms of feasibility and likelihood. The impact of some of these attacks may be profound.
Generalizing this,
Types of potential attacks involving the hostile use of GMOs. The individual threats are roughly ordered from bottom to top in terms of increasing risk-potential—which correlates with the difficulties attackers are facing in effectively realizing those attacks. The impact is also hierarchical. Risks at the lower level are inherited at higher-level attacks. For the feasibility of the individual attacks and further discussions, see sections 4.2, 4.3, and 4.4. HEGAA, Horizontal environmental genetic alteration agents.
This section lists a variety of effects that attackers may be aiming at. These are summarized in
Potential direct targets and attack aims.
An attacker could rely on traditional (e.g., |
(a) Designed spreading of toxins and harmful substances to targeted hosts (specific targets in the food chain). |
Plants have innate players and mechanisms (toxins) to defend themselves against threats like pests and pathogens. The ability to produce toxins is need-based (some toxic plant pathways are inactive). An attacker could |
• Increased chemical toxicity from plants into feed and food products; tailor-made increase of existing (weak) mechanisms that make plants pathogenic to consumption (e.g., possible increased toxicity in additional plant parts and/or in various stages of the growing/harvesting/processing cycle). |
Specific promoters can not only upregulate the expression level but also lead to differential expression of transgenes in specific tissues (Arpaia et al., |
• This may allow the (clandestine) introduction of toxins and harmful products, disguised as popular plant parts used for food and feed. |
Targeted interference of cellular pathways, leading to an upregulation of immune reactions in those consuming the plant. | Disruption of immune response; induced hypersensitivity response to certain nutrients in animals and humans. |
Potential off-target and indirect attacks.
The intended deletion or silencing of genes or mechanisms with plant protective properties. | The disruption of protective mechanisms (e.g., reduced levels of secondary metabolites, Arpaia et al., |
Attacks in form of under-appreciated relationships and off-target effects. For instance, Bt crops express specific Cry proteins within the plant. The mode of action relies on interactions with specific midgut proteins of the targeted insect pest. An attacker may heighten or broaden the toxicity of Cry protein variants to interfere with new targets (see also |
The disruption of mechanisms involving off-target species (e.g., the human microbiome or viome), their relationships and synergistic effects. |
An attacker may increase dominance of a new trait (e.g., through gene drives), or through misuse of infectious genetically modified viruses and other horizontal environmental genetic alteration agents, see Reeves et al. ( |
Increased susceptibility to pests and pathogens; harmful effects on non-target-organisms; biodiversity disturbance, species displacement, and extinction; disturbance in soil micro-environment and species of ecological concern. |
As known structures, including mobilization genes or origin sequences, are routinely screened in an attempt to ensure the safety of engineered products, an attacker may try to find unrecognized vulnerabilities often in form of new insights derived from GMO research.
It is possible that some of the most promising methods for supporting the utilization of GMOs may become a basis for attacks.
For instance,
Bt toxins.
To delay evolution of pest resistance to transgenic crops producing insecticidal proteins from |
|
A new method to combat Bt toxin resistance was recently proposed by Badran et al. ( |
|
A very effective way for delaying the evolution of pest resistance to transgenic crops is the “refuge-in-a-bag” approach, which consists of a random mixture of seeds of Bt and non-Bt plants of the same crop. | An obvious form of attack would consist in creating a different mix of seeds. For instance, as cross-resistance is expected to be stronger between toxins that are more similar (Carrière et al., |
Further, possible dual-use potentials of CRISPR/Cas technologies are listed in
Possible dual-use potentials of RNA-guided CRISPR-Cas9 systems to target plants.
An attacker can direct Cas9 to cleave sequences involved in basic plant housekeeping mechanisms. | Weakening of plants, manifestation of disease, interference in plant interactions with other biota |
Such a“miRNA stacking attack” may be targeted to deactivate the expression of critical genes in plants, which may be harmful to the plant itself and disrupt plant interactions with microbes, insects, and herbivores. | |
• In contrast to insect herbivory, the breakdown of certain plant secondary metabolites (glucosinolates) produced during the processing of oilseed meal have a harmful effect on animal thyroid function. The use of animal feed containing these glucosinolates has a negative effect on animal nutrition because of their goitrogenic properties (Borgen et al., |
|
Apropos of indirect defense and the expression of volatile substances that attract the natural enemies of plant herbivores. This is exemplified via a specific case-analysis. |
A number of detrimental effects of weakened expression of volatile substances—which would be the point of attack—are described by Pechanova and Pechan ( |
Finally, possible dual-use potentials of gene-drives
Possible dual-use potentials of gene-drives.
Gene drives were originally invented for the alteration of sexually producing wild populations (Esvelt et al., |
The consequences of a gene-drive escaped into the wild are generally believed to be profound; or catastrophic, as gene-drives enable the spread of various traits, including malignant ones, such as disrupted defense mechanisms and resistance to various herbicides. |
According to their original design (involving pest/pathogen populations in the wild), the intent of gene-drives is to drive a new trait through entire populations. The significance is that all the offspring of a gene drive organism that mates with, or is pollinated by, a natural organism will have the driven trait. Nonetheless, to affect a difference it is necessary that the drive can make its way through sufficiently many generations. Therefore, the impact of (manipulated) gene-drives in the context of cultivated crop plants seems minimal.
However, a new concern may arise when introducing a gene drive into a situation where plants are able to form self-sustaining populations, such as in non-GMO environments (see also section 4.7 below). This way, the attack vectors would not be commercial GM seeds, but in fact natural crop plants. Although the exact mechanism and impact of such attacks are open to debate, the consequences may be catastrophic.
Traditionally, security issues have been captured by cryptographic techniques which then expanded into various branches of cyber-security and information theory. While the biologic realm is also interested in (biologic) “information,” the latter involves many additional features than the one modeled by the cyber domain. Nonetheless, it seems beneficial to parallel possible motivations driving the threats within the cyber and the biologic domain. This is depicted in
On Motivations and Intent.
Gain unauthorized access to information, in order to intimidate or coerce a government or its people in furtherance of political and social objectives. | Gain unauthorized access to power in form of blackmailing. A perpetrator may claim to have mingled manipulated seeds into the legitimate supply chain, and threatens to effect their targeted (or large-scale) release and distribution. | Low. |
Impersonate another user or product for the purpose of: |
Impersonate a developing company resp. a certified GMO for the purpose of: |
High. |
Impersonate another user or product for the purpose of: |
Analogous. | High. |
To cripple critical targets (based on political, social or religious objectives). | Diseases inflicted upon plants to harm the economy (or a competitor), to threaten the food supply, and pose health hazards on humans and animals (possibly with racist intentions). | High. |
• Fraudulently restrict the license of others. |
Introduce (a) illegal, or (b) harmful features into certified GMOs to evoke legal actions against biotechology companies and to bring certain producers or companies into discredit. | Medium (a), high (b). |
Undermine confidence in a protocol or service by causing apparent failures in the system. | Undermine confidence in GMO production, sales, and politics. | Unpredictable. |
Insider attacks (revenge, personal gain, personal or political motivations). | Analogous. This may include, (a) the release of inadequately performing GMOs during testing phases or trial-and-error experiments, or (b) the illicit release of those that have been manipulated in clandestine. | Low (a), high (b). |
The challenge of breaking something that is believed to be secure; exploiting the naivety and ignorance of those susceptible to intrusion; being the first to demonstrate that attacks or security breaches are possible. | Analogous. | High. |
Infliction of harm, (a) directly or, (b) with the aim of creating fear and shock). | Analogous. | Low (b), high (a). |
It should be stressed that the cryptographic view is radically different than the one prevalent in the biological/medical sciences. In the latter case, it is about saving and supporting life. It is difficult for those engaged in those disciplines to appreciate a kind of opposite driving force, such as the challenge of breaking a system. Anderson (
Concern has already been raised before regarding the—minimal—abilities required for modern synthetic biology technologies to be exploited in a malignant way. Frazar et al. (
Furthermore, already a decade ago, Graham and Talent (
The implications are sobering. Dunlap and Pauwels (
Beyond doubt, the CRISPR-Cas system is a most effective gene-editing tool, and it is readily available. Although the system is targeting individual cells, it has been used to edit various organisms or cells from organisms. With respect to plants, editing efficiencies of the first CRISPR-Cas9 technologies for plant editing initially were relatively low (Mao et al.,
One of the key barriers to achieve noticeable effects lies in the delivery of the genome engineering reagents into plant cells. As plants are complex multicellular organisms, attackers need to find a way to manipulate sufficiently many cells to achieve the highest level of harmful effects (see
Although practically this can be realized via Agrobacterium-mediated transformation methods or alternatives (see e.g., Globus and Qimron,
To illustrate this point, a rather benign incident was described by Peccoud et al. (
The GMO production pipeline may harbor similar vulnerabilities. What an attacker can exploit is that a (digital) description of a product does not have to be the same as the product itself. As a result, skilled attackers could manipulate the underlying automation and digitization procedures to clandestinely tamper with authentic descriptions of the constructs encoding the CRISPR-Cas9 components.
The switching of authentic with fabricated genome engineering instructions is a serious threat (see also “Mapping the Cyberbiosecurity Enterprise” as a recent Frontiers Research Topic) and may have contributed to a critical feed contamination incident in the EU. In July 2014, Germany detected a viable genetically modified
Correspondence between German diplomats, Chinese authorities, and the manufacturing company confirmed that there were critical genetic differences between the strains the company asserted to be using and those detected in Germany. Paracchini et al. (
Vulnerabilities arising at the interface of the digital and physical realm have hitherto received insufficient attention. They may indeed have played a role in the B2 feed contamination. Whether or not they can fully explain this incident, their potential for tampering with the GMO production pipeline cannot be ignored.
In order to induce any observable health effects among consumers (
As previous incidents involving the import of unauthorized GMOs have shown, the potential of commingling manipulated seeds along similar routes must be taken seriously. Additionally, there is the risk about the mingling of seed from laboratories or field trials (see also section 4.3). Holst-Jensen et al. (
As our mechanistic understanding of plant gene function and regulation is rather limited (see e.g., Rhee and Mutwil,
The significance of this approach for attackers lies in the “error” outputs of the method. It is known that loss-of-function mutations disrupting genes may have extreme effects on the phenotype and also that the editing of regulatory elements can have unpredictable results (Rodríguez-Leal et al.,
In spite of tremendous progress since the introduction of the CRISPR/Cas technologies just a few years ago, the simplest types of plant editing involves loss-of-function mutations into genes (Scheben et al.,
Thus, as predicting the phenotypic consequences of a specific mutation
To the best of my knowledge, the types of risks introduced in this paper have not been described before. The lack of familiarity with new technologies presents a challenge to risk analysts who wish to precisely identify the hazards and the likelihood and magnitude of these outcomes. “The Regulation of Synthetic Biology—a Guide to United States and European Union Regulations, Rules and Guidelines” (Bar-Yam et al.,
While new technologies have been leading to important insights how to advance plant breeding and agricultural issues, these may establish unrecognized risk potentials.
Attackers may pursue various levels of attacks (see
It should be noted, though, that attackers would pursue opposite goals than those pursued during legitimate GMO production. The latter is concerned with biocompatibility issues and regards unintended risks to health and environment. These practices are aimed at ensuring that the new traits get introduced without leading to unexpected effects or harming non-target organisms. Moreover, a great challenge in producing CRISPR-based GMOs lies in poor CRISPR-Cas specificity leading to frequent off-target editing. This must be avoided if regulatory approval is required and considerable effort has been devoted to alleviate this problem. Such an orientation is singly minded toward the achievement of very precise goals—targeted and critical phenotype expression levels of the intended modification. Everything else other than the specific goals would be unacceptable.
Yet, attackers would be dealing with the opposite scenario. Possibly, they do not even want a (targeted) effect (see
Even if the clandestinely introduced alterations did not lead to the anticipated linear causal relationships (i.e., the intended harm), they could have unexpected effects that are detrimental nonetheless. These negative effects may involve the plant itself, but could be on a more general level, from cellular to political and economic. A critical factor here is public opinion. According to Parrott (
The very announcement that one, or possibly many, manipulations have happened, may lead to considerable unease and angst. Attackers could further stir the fear of the public by claiming that some of the introduced modifications are intentionally disguised within the genome, or that some of the biochemical pathways to introduce toxic effects are clinically significant but difficult to diagnose. Take the example of chemical toxicity described in
Now add to this the problem of analytically identifying the purported hazardous gene edit. New breeds of GM crops have led to substantial debates regarding their biosafety, commercial use, and regulation. The very fact that attacks through and on plants (possibly extending to non-GMOs, see section 4.7) are difficult to detect would yet further increase public unease and increase doubt in these new technologies and the companies producing them.
As there is limited comprehension about many biocrime risk potentials, this may be exploited by those seeking to do harm. Dunlap and Pauwels (
Depending on their goals (
The challenge of rapid and efficient identification of manipulated GMOs could additionally be exploited via blackmailing attacks. In this case, perpetrators would just need to provide sufficient evidence regarding the plausibility of the attack. In order to do that, attackers would only need to make sure that some modifications get detected, e.g., via genetic or chemical analysis. Such threats cannot be taken lightly. Given that both the authentic and the manipulated GMOs would carry the same DNA signature elements, the extent of the intrusion—from targeted and isolated manipulations to sabotaging the manufacturing company—would not immediately be clear.
Concerns about GMO have had a significant impact on public opinion and disputes over unauthorized GMOs are known to have considerable effect on economy and trade (Holst-Jensen et al.,
While modern technologies may pave the way for various forms of biocrime involving plants, the potential of counterfeiting makes GMOs even more vulnerable. The critical point is how these attacks could get detected. If manipulations could easily be identified, then they would be less “attractive.” Here is where the predicament of counterfeiting comes in. For decades, the problem of GMO authentication seemed to be solved, and methods utilizing the unique flanking sequences were believed to be most reliable. Unfortunately, due to the advancement of modern gene editors and technologies, this is no longer case.
A much easier way to introduce harm than discussed herein would consist in genetic manipulations to develop more virulent or resistant pests/pathogens in order to harm plants or their microbiota. When introduced in form of manipulated gene-drives, this would have a large-scale detrimental impact on plants as well as non-target organisms and have a cascading effect on the entire food-web. At present, the potential for manipulating gene-drives is difficult to assess.
The reason that GMOs are considered in this study lies in the identification issues. By clandestinely exchanging GMOs with adulterated versions, and by actually utilizing and verifying the authentic identifiers, an evil-doer appears in sheep's clothes. Although protocols utilizing flanking sequences are particularly vulnerable, the same is true for other methods. As long as there is no easy and rapid way to authenticate GMOs, this may enhance the scaled-up distribution of hazardous counterfeits.
As this is the first study of this kind, the focus was mainly on GM plants. The full extent of biocrime vulnerabilities involving biological mediums, including animals and viruses, needs to be further analyzed (see, e.g., Reeves et al.,
Skilled attackers might be able to exploit numerous insights about functional relationships, synergistic effects and emergent properties, plasticity of gene expression and memory effects, or under-appreciated contributors such as the active role of proteins or metabolism, the human/plant microbiota, and epigenetic regulation. The list seems endless. Nonetheless, without a mechanism to scale-up the intrusion, many attacks might only have a rather limited—that is, local—impact. The discovery of gene drives (Esvelt et al.,
Although biosecurity issues of gene drives have been the subject of several studies (e.g., Oye et al.,
Related questions arise relative to horizontal gene transfer (HGT). While the frequency of (unintended) HGT has been under debate, certain transgene features are known to increase introgression into wild counterparts (Tsatsakis et al.,
Apropos of GMO authentication. One of the most critical questions would be to develop a rapid and effective GMO method that, (1) provide a signature type function which, (2) cannot be copied or transferred, and which (3) detects arbitrary manipulations that might be hidden inside the GMO. Given that GMOs are not a “closed box” it seems that such signatures would have to rely on something different than just the genomic sequence information. Extending the signature paradigm to the various omics settings seems like a daunting task, especially in light of the plasticity and flexibility that is governing natural processes.
This article suggests that criminals may exploit modern gene editing technologies via new forms of attacks that involve the clandestine manipulation of market GMOs (or the malicious insertion of genetic modifications into ostensibly unmodified plants). The identification and analysis of the harmful genetic manipulation of plants as an attack vector shows that these concerns need to be taken seriously, raising the prospect not only of direct harm, but of the more likely effects in generating public concern, reputational harm of agricultural biotechnology companies, law-suits, and increased import bans of certain plants or their derived products.
If an agent can manipulate legitimate GMOs, the emergence of undesirable traits amongst GM plants and their derived products might not immediately point to an attack. Disguised as a certified product, researchers might look for other reasons to explain such effects, trying to find reasons via unrecognized natural phenomena or biochemical pathways. The investigations might point away from the reality of an attack, let alone point to the true identity of the attacker. It might take a while until the DNA of the harmed plants would become the target of investigation, and when, say, full-genome sequencing would point to the presence of the foreign gene(s) (if existent) responsible for the observed, new, harmful traits. However, attacks based on more recent technologies that influence the proteome or metabolome would be much harder to detect. Similarly, the effectiveness of authentication and surveillance protocols may be overridden by skilled misuse of newer technologies such as gene-drives, phage assisted methods, or
While some of the scenarios described may be a bit far-fetched, this was done with the intent to anticipate a large range of approaches that potential attackers might be pursuing. This paper should be seen as an invitation to start a dialog and to raise awareness. It is the type of study that never realized would enjoy it's greatest success.
The author confirms being the sole contributor of this work and has approved it for publication.
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
After having to battle severe illness, this is the first paper that I wrote in several years. This would not have been possible without the help of the editor and the many insightful comments by the reviewers. I am deeply indebted to their incredible support. This has been the most supportive and helpful feedback I have ever received when writing a scientific work.
1The Centers for Disease Control and Prevention (CDC) is the leading national public health institute of the United States,
2for simplicity, herein referred to as GMOs.
3Throughout, “harmful GMOs” are meant as a result of attacks that deliberately introduce harmful features.
4I am indebted to the reviewer who suggested to include gene-drives into the present analysis.