IEEE Technology and Society Magazine - December 2020 - 33

seems that the existing attack vectors indeed pose a
threat to the cybersecurity of the technology.
Looking at the farmers using the technology, understanding what they do to ensure their cybersecurity, and
more importantly, whether they share the same attitude
towards data loss being critical, paints a more nuanced
picture. In interviews with users of the sensor technologies, we found that they were not as worried about data
loss, whether data itself leaking by accident or being
maliciously extracted. Furthermore, the farmer mentioned that to their knowledge other colleagues in the
field did not worry about this either, even though they
admitted to having a central local server that holds all
their real-time and historical data. When asked why they
did not consider data loss to be a threat, they noted two
key things: unlike the vendor, they are not worried about
loss of real-time data, as this is used for day-to-day operations rather than long-term analysis and decision making; and they simply do not consider the data to be
commercially sensitive. In fact, they shared it freely: " if a
veterinarian calls and asks for the data, they get it. If a
researcher calls and asks for the data, they get it. "
Farmers did note the impact that a loss of real-time
data would have on their ability to immediately react to
their cows, but, besides having never experienced such
data loss before, they noted they knew how to run their
farm, and that years of experience in day-to-day work
meant that they perceived the impact of not having this
data on the productivity and welfare of their livestock as
negligible. Regarding historical, or aggregated data, they
explained that loss of such data is certainly an issue
that has to be dealt with, as this data informs their decision-making. However, dealing with it was considered
trivial because relevant data could simply be obtained
from colleagues, as there is an existing culture of being
open and sharing such data.
They perceived this attitude towards data sharing to
be grounded in the history of the agriculture sector in
Israel, where farming has effectively descended from
kibbutzim [singular-kibbutz -]: ideological collective
communities hoping to spread socialism and equality
" to all corners of the land " [22]. These communities
were primarily centered around agriculture, and built
on a utopian ideology of social responsibility and sharing of labor [31]. Over time, and in part due to economic crisis as well as increasing individualism, many
agricultural k ibbutzim became privatized, and
detached from the original utopian visions, but, nonetheless, their history had a lasting effect on the attitude
of people working in the sector. Thus, while in modern
times the agriculture sector is no longer exclusively
built on collective farming models, the open attitude
of the people involved in the agriculture sector has
prevailed. Shared responsibility is effectively seen as
DECEMBER 2020

∕

shared benefits, especially when it comes to data analysis. Dairy farms provide data to a central authority,
which allows for more insights than any individual
farmer could generate on their own with their livestock. Almost all dairy farms in the country participate
in this system, and benefit from the research and analysis performed centrally to understand their own cows
better. As such, they are happy to share data, and do
not consider it a threat if other farmers learn details of
their operations (whether livestock data or even financial data), even meeting up yearly to compare and
learn from each other.

Farmers said that, due to their years
of experience in day-to-day work,
the impact of any loss of data on
the productivity and welfare of their
livestock would be negligible.
Instead, farmers' primary concerns in terms of
potential cyberattacks were not focused on access to
data held in their farm's IT layer per se, but rather on
the accuracy of the data captured by the operational
technology, the actual physical sensors. They noted that
it would be a major problem if sensors were somehow
compromised and subsequently provide erroneous data
without them realizing, as this would lead to making
decisions on faulty data, potentially impacting the productivity and welfare of the livestock. However, they perceived this as very unlikely, assuming it to require
physical access, and relevant know-how of the hardware
itself. Indeed, this kind of cyberattack is unlikely given
the known threat actors in this context, as corrupting
data would decrease its potential value for anyone
involved. Commercial sabotage could thus be a main
driver for such attacks, although in this socio-cultural
context such attacks would not make sense as farmers
freely share data and collaborate, rather than compete.
Given the dairy farmers' propensity for sharing files,
and learning from each other, this thus seems an
unlikely threat, as it only reduces potential value for
any actors involved (less worthwhile data to learn from,
whereas affected target can still learn from untampered data shared by others). Another critical aspect
supporting the cybersecurity of the technology is that
farmers in this culture are open to the vendor gaining
access to their information technology in order to configure the operational technology. That is, they happily

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

33
IEEE Technology and Society Magazine - December 2020

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - December 2020
