IEEE Technology and Society Magazine - Winter 2014 - 76

be a continuous process of invited
enquiry through purposeful organized
public engagement before during and
after the development of IoT. A similar need for this type of engagement
was already discussed by Albrecht and
McIntyre in [11] for the risk of privacy
threats in relation to the use of RFID
technology. In the domain of ethics of
ICT in general, and of IoT in particular, we do not have the luxury of the lab
experiment to enquire about the deep
transformations that these technologies may bring to individuals, societies, and the planet in general.
As in other areas of operation,
with ICT those transformations are
not necessarily material, but rather
of a profound or nuanced normative,
and even constitutional, nature [12].
Also, as with so many other technologies, we are not able to anticipate the
impacts of technology at all levels -
including differences according to
context - in order to govern them.
Therefore the responsibility of these
transformational developments needs
to be debated and shared. Rather, discussions of values, norms, and technology appropriation should be part
of collective deliberation. But probably, the first question we need to ask is
why are we doing what we are doing?
How many and what aspects of our
own agency, ability to think, feel,
experience and decide are we willing to sacrifice? Where and how do
we set the boundary between living
and merely functioning? Do we need
to preserve human agency and experience as we know them or witness,
more radically, the possible transformation of their very nature ? [3].

Challenges to Building
a Human-IoT Trust
Relationship
In the previous section, we investigated the drivers and the need to
support trusted relationships and
agency in IoT. In this section, we
identify the major challenges to
support these objectives.
One significant challenge is the
variety of technologies and types of
devices that will be used in future IoT
76

infrastructures. Not only will many
IoT devices have resource constraints
(e.g. processing power, memory), but
there will be a potentially unbounded
number of interacting entities, with
substantial differences in the interaction patterns [13]. Traditional security solutions like Role Based Access
Control (RBAC) and Attribute Based
Access Control (ABAC) systems may
not be adequate to support the heterogeneity and complex distributed environment of IoT. For example, RBAC
and ABAC make hard to enforce the
least privilege principle, as discussed
in [14], and have very limited expressiveness considering complex policy
rules including context-aware obligations with temporal constraints,
which are common in IoT scenarios.
Techniques to implement security and trust in IoT systems can be
defined for a specific context, but may
behave in an incorrect way in a different context. For example, a trustful IoT framework designed for the
office context where an IoT device
operates in a controlled environment
may be vulnerable when it interacts
with non-trusted IoT devices in a
public space. Context-awareness is
therefore a required characteristic of
a possible security and trust solution
to be adopted for IoT.
IoT devices will also be vulnerable
to classical hardware failures, which
need to be monitored and managed in
order to ensure a trustful relationship
across the IoT system. An integrated
risk assessment approach is necessary to address these aspects and
also malicious activities and attacks,
which may lead to the problem of
false information. One approach to
solve these "uncertainties" is to adopt
trust management mechanisms, similarly to what was already adopted for
P2P and grid systems, i.e., exploiting
the collaboration between nodes and
assigning reputation and trust values,
which have been proposed by [15].
Another important aspect is identification. Different forms of identification (numbering, addressing and
naming) are key components to provide trust in IoT, provided that the

identification mechanism is tailored to
the needs of trust on the system and do
not result in a lack of privacy. Unfortunately, many of the existing naming,
numbering and addressing schemes
have been designed to address specific requirements and there is not an
agreed abstract approach for identification which can be used directly
in concrete IoT systems. In addition,
each of these schemes has its own
established identification hierarchies, which may contrast with the
unbounded number of interacting
entities and interaction patterns mentioned above. In most scenarios, the
identification of the IoT device is also
linked to its status (e.g. active, dormant, de-activated or retired from the
market), which is difficult to maintain
in a distributed and dynamic environment as the IoT is expected to be.
Many of the challenges above
have been already faced for the Internet and, although some specific technical solutions could be borrowed, a
considerable effort to adapt them to
IoT is needed. For example, trusted
execution environments (TEEs)
are already adopted as secure areas
for code and data in the processor
of smartphones and other mobile
devices, but similar solutions for
more constrained IoT devices are
necessary. Therefore, because of all
these challenges and related efforts,
we cannot reasonably expect that IoT
systems and devices will be able to
be trustworthy in any condition and
context, but it would be useful to
have a degree of confidence in the
level of trust, which can be offered
by the IoT towards the user. Even if
the adoption of common standards
can mitigate many compatibility
problems due to heterogeneity of
devices and applications, we argue
that an integrated security framework including risk analysis, identity
management, trust relationships, and
technical security policies may represent an alternative solution to these
problems. Indeed, the combination
of technical solutions embedded by
design in the systems and the preservation and enhancement of places for

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

WINTER 2014

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - Winter 2014