Table 3.

Examples of “insecure channels” in the field of cyberbiosecurity.

Insecure channel	“The message” (what is to be secured)	Feature of a secure-able channel	Comment
DNA replication: The process of passing on a parental piece of DNA to offspring	The specific DNA sequence	The DNA sequence is the same before and after replication	Numerous cellular repair mechanism turn the potentially insecure/noisy channel into one that is secured
Artificial plasmids. These are carefully designed to lead to a specific trait. Specifics of the expressed phenotype are coded in the artificial sequences	The artificial DNA cassette	The sequence information of the artificial construct is the same, regardless of the lab or environment that it is utilized by. To be “secure-able” means that this information can be traced back to its original/intended sequence	Sequencing of the plasmid allows to reveal its complete and detailed sequencing information. While this is costly and technically demanding, this shows if the channel (the sequence information encoded by the plasmid) matches the expected sequence [as e.g., can be verified by secured databases (see Peccoud et al., 2011, 2018; Wilson et al., 2012; Dunlap and Pauwels, 2017)]
Raw data, health related information, medical databases (storage of man-made information, as opposed to sequence information in living organisms)	The digital information about medical insights, health records, etc	The digital data remain unaltered (same information regardless of when and by whom it is read out), accessible only to legitimate authorities, and whenever needed	Once the information is in place, this essentially is a cyber-problem and can therefore benefit from existing cyber-related tools
Artificial DNA sequences, DNA as information storage	The message is the information to be stored in form of artificial DNA bases	As above	Need to filter out alterations due to DNA processing. Can benefit from alignment-based methods such as distance-measures (e.g., Federhen et al., 2016) and additional coding-theory and cyber-based tools to identify, correct, or minimize any errors (see e.g., Mueller et al., 2015)
Expression of a transgene via a GMO. Targeted phenotypic trait and expression levels	“The message” is the specific transgene. The channel that aims to be protected is the transgene only “The message” is the entire genome. The channel that aims to be protected is the entire organism	The transgene achieves its targeted phenotypic expression, relative to its trait, expression level, and in the context of its intended (molecular, biologic, cellular) environment	The phenotypic expression can be influenced by illicit genetic modifications outside of the transgene. If integrity is verified with respect to the transgene only, such covertly introduced modifications are not detected. They lie outside the specific channel To obtain a secure-able channel, it needs to be the case that (1) The entire genome can be sequenced, (2) The sequencing information obtained in different contexts and circumstances always lead to the exact same sequence (possibly including predictable differences within a certain range or distance)
Modern gene-edited plants and crops (see e.g., Grohmann et al., 2019)	Unclear what the message is. This is because the intended effect is based on a range of expression levels via specific biochemical pathways, which are dependent on their context and environment (here, environment is meant across the full spectrum, from molecular to gross)	The intended outcome is a spectrum of traits, depending on the specific context and environment. Here, secure-able would mean the same spectrum of phenotypic expression, as informed by different, discrete conditions in a clearly causative way	It seems much more difficult to secure a channel like this, where there is no tangible fixed, physical message that can be identified as the key information to be protected

The key feature of insecure channels often can be formulated in terms of existing cryptographic primitives. For instance, all channels involve attributes that aim at leaving some information unaltered (integrity). Insecure channels in the cyber domain build on the salient feature that these can in fact be “secured.” In the context of integrity this would mean that the original intended information can be recaptured. In cryptography, what needs to be secured is typically called “the message.” It is important to note that this term has nothing to do with our contemporary usage of this word. Here, it describes the defining characteristics of the insecure channel. By identifying “the message” involving biological mediums it is found that many of the insecure channels are in fact “insecureable”.