3.2.1. Innovation and classic forms of responsibility

A first approach to responsibility in innovation may be developed on the basis of contracts between an organization and its employees, suppliers or shareholders, as presented in Chapter 1 in the more general context of business; these contracts provide the framework for the legal existence of companies, whether or not they constitute a source of innovation. They define a field of legal responsibility, as discussed previously, which may also be applied to innovation. The actors involved in innovation, from start-ups to small- and medium-sized businesses, big companies to researchers and research institutions, project incubators, competition centers and public authorities, are all subject to forms of responsibility governed by law and by the contractual relations between these players.

Thus, in cases of breach of contract or failure to respect the obligations set out in a patent, or when the safety standards to be applied when manufacturing a new product are neglected, for example, a company may be liable to a fine or criminal prosecution, according to the nature of the violation or offense (see Chapter 1, section 1.2). What was true for business in general is true for innovation: legal regulation, based on the question of responsibility, is essential. For example, laws against workplace discrimination, or the creation of thresholds for greenhouse gas emissions, have observable and measureable effects on the decisions of social actors. These standards have a limiting power; they condition current and future actions, and enable the pursuit or achievement of goals that have been identified as morally desirable (in this case, workplace equality and reductions in pollution).

However, the conclusion set out in section 1.2.2. for companies in general is also applicable to innovation: legal responsibility is not sufficient to cover the full field of action in innovation, as it only relates to responsibility insofar as damage has been caused; it does not lead to a normative commitment from social actors, and does not always result in proactive approaches [PEL 16a].

Furthermore, legislative developments often lag behind technological evolutions [GOO 83, MEL 08, ROU 11]. This limitation is even more striking in the context of innovation. In addition to the different time scales involved in technological and legal developments, the definition of an observable and measurable consensual aim, required for regulation, is not always easy in the case of rapidly evolving innovative technologies. The debates concerning the regulation of nanotechnology in Europe are a case in point. The current REACH⁷ measure, relating to the recording, evaluation, authorization, and restriction of chemical substances within Europe, contains no specific provisions for nanotechnologies. This regulation requires companies to make declarations concerning the substances that they produce or market, and to ensure that they do not have a negative effect on human health or on the environment. In practice, however, a number of nanomaterials have been released onto the market without prior registration or monitoring. As REACH only requires companies to specify the composition of chemical substances, there are no automatic obligations concerning the specific characteristics of certain nanotechnologies in terms of size, structure or new properties⁸. A number of actors (including the European Parliament, national health agencies, non-profits and actors in civil society) have condemned the difficulty of carrying out sufficient risk evaluations in cases where observable substances have not been declared, and have indicated the need for specific regulation in the case of nanotechnologies. This debate highlights the fact that a clear, measurable and consensual social aim – in this case, regulation of the industrial development of nanotechnologies – is not always an attainable goal; nor is it immune to strategic positioning, as industrial actors are unlikely to want specific regulations that impose additional limitations on their activities. The issue of regulation is therefore a complex one.

In this context, as we have seen, CSR is a source of norms that are complementary to and aim to compensate for the shortcomings of legal provision. However, while the field of responsibility in question includes a number of dimensions and business activities, as seen in theoretical publications and in international standards such as ISO 26000⁹, it is not specifically intended to respond to the challenges of innovation. It is thus necessary to identify practices and theoretical proposals from CSR, which may be of use in developing an approach to responsible innovation, and to consider the limitations inherent in applying CSR principles to this complex domain.

3.2.2. CSR and innovation: scope and limitations

The different principles recommended in the context of CSR (as presented in Chapter 1) provide a primary theoretical framework for the notion of responsibility for innovative companies. Taking account of company stakeholders and their interests, along with the importance of dialog with those who are affected by and who affect company activities, creates a set of general recommendations, irrespective of whether the justification theories used are based on deontological approaches to morality or on consequentialist reasoning. This set of recommendations, from which local and concrete standards may be drawn, allow us to ensure that scientific and technological development are carried out in a pluralistic manner, i.e. taking account of the potentially conflicting interests of different social actors.

However, specific problems relating to company size, influence, means (which may be limited) and to the radical breaks, which may be brought about by innovation, make it difficult to transpose CSR principles as-is, wholesale, to the context of innovation. The following analysis of problems that are specific to innovation, and of the way in which the CSR framework is able to provide a more or less appropriate response, aims to highlight the essential dimensions that must be taken into account in developing principles of governance for responsible innovation.

3.3.1. Epistemological and moral innovation

Models are not neutral. This question has long been considered in the field of epistemology, the analysis of theoretical frameworks of thought. Alongside other concepts, such as scientific truth and argument, the concept of innovation corresponds to a social construct that has developed over time, and which is partially dependent on the theoretical frameworks used to understand it. The different definitions of innovation, or the distinctions made between disruptive and incremental innovation (see section 3.4), constitute theoretical attempts to rationalize and act upon reality. Theoretical models thus act, at an initial level, by highlighting, distinguishing, sorting and ordering perceived reality, and identifying elements of this reality that need to be assembled, connected to other elements, or left aside. On a second level, models thus have a feedback effect on the reality to which they are applied: once implemented, and before being destabilized by radical changes in perspective, a theoretical model determines the nature of questions to be studied and the methods used in building a response [KUH 62].

There is thus a form of epistemological responsibility pertaining to the different actors who create these models, as the effects of models on the reality that they aim to represent may be dramatic, in cases where hypotheses and design choices are not openly disclosed and discussed. Much can be learned from the example of the financial crisis of 2007–2008 in this respect; it highlights the fact that the theoretical frameworks used to understand innovation are themselves associated with an initial level of responsibility, as they determine elements that are to be considered problematic (and thus which require an innovative response), the types of solution proposed, and, more generally, the way in which certain effects may be considered to be beneficial or otherwise.

3.3.2. The financial crisis and the responsibility of theorists

Following the rapid rise in popularity in certain innovative products involving poorly assessed risks (subprime mortgages and derivatives based on these low-security loans), the international financial system experienced a large-scale global crisis, starting in the United States in summer of 2007. The crisis rapidly spread around the world, notably resulting in a public debt crisis in Europe. Without offering a detailed map of the responsibilities involved in this crisis, it would appear that commercial banks, central banks, credit institutions, rating agencies, governments, economic models and financial models have each, in turn, been held responsible for the emergence of this crisis, its scale and its propagation.

The responsibility of the scientific community in this context is not the most obvious element nor was it one of the first to be recognized. However, the theoretical models put forward by financial and economic analysts had a particular part to play through their status as the social “technology” used in the identification and understanding of the risks involved in the financial innovations linked to subprimes. Actor behaviors in this crisis were broadly determined by a set of beliefs and theoretical models (both macro- and microeconomic), i.e. by a set of knowledge circulating between different academic communities and the decision makers, regulators, bankers and other practitioners involved in the financial world. This example is a clear illustration of the fact that models are not neutral, which has implications for those who propose and defend them [DOU 03, PID 07]. At a theoretical level, in representing a performative conception of the world in the field of economics, the choice of a model generates responsibility. This does not only apply to researchers; in this case, the financial actors who applied these choices also have their share of responsibility.

In his work on the role of economists in the 2007–2008 financial crisis, Cartapanis [CAR 11] notes that accusing economists of failure in predicting the crisis would be a mistake. Even the most sophisticated models are not always able to offer clear predictions concerning tipping points and the emergence of a crisis of confidence. However, the author denounces the epistemological and political consensus that has emerged in favor of market deregularization, and the theoretical disinterest of different economic and political actors with regard to a situation in the United States that was highly characteristic of past crises: high deficits in the balance of payments, high levels of household debt and speculative bubbles in the assets market, encouraged by relatively low interest rates set by the federal reserve, and uncontrolled use of financial derivatives. Cartapanis states:

“The risk of a systemic crisis was dramatically ignored by the mainstream¹⁵, not because the phenomenon was improbable, but simply because the conjecture was impossible within the framework of the models used by a large sector of the profession. […] The dominant streams of market finance and financial microeconomics thus legitimized a whole set of hypotheses relating to the stability of financial innovations and the harmless nature of various macroeconomic imbalances […] Among both financial intermediaries and decision makers, many individuals, operating at the interface of theoretical reflection and action, drew a political message from this – neglecting the local and specific character of the obtained results, which had been highlighted by academics, even within the mainstream – justifying financial deregulation, the use of fair-value type accounting standards, the involvement of banks in intermediation and market activities, the development of innovations, and trust in the capacity of markets to self-regulate” [CAR 11, pp. 152–154].

These different assumptions fed into the idea that market regulation was not necessary. They also resulted in wide acceptance of the excess of macroeconomic fundamentals in the United States (balance of payments, a real estate bubble, financial innovations linked to credit derivatives, etc.); in different theoretical models (for example Keynesian-inspired approaches), these elements would have raised alarm bells at an earlier date, potentially leading to corrective actions and avoidance of the crisis.

Taking a similar argumentative approach, Walter and Pracontal [WAL 09] highlight the hegemony of risk models based on a mathematical understanding, with its origins in the English-speaking world, which considers that markets behave following the Brownian laws of motion. According to the authors, the dominance of this model among economists and financial analysts prevented any other mathematical models from emerging; these models might have had a greater capacity to predict crises, notably in the case of 2007–2008. In their view, the trust placed in the Brownian-type model by most actors in the financial sphere is not always scientifically validated. The financial innovations resulting, in part, from these models – another form of social “technology” with potentially significant consequences – are poorly suited to the management of contemporary crises. In this case, a relatively unchallenged epistemological choice prevented alternative risk analyses from emerging, with clear effects on the emergence of financial products.

This brief consideration of the role of economic and financial models in the 2007–2008 crisis is useful for our examination of responsible innovation. Actors (credit institutions, central banks, investment banks, financial analysts, academic researchers, etc.) are responsible for the models which they help to produce, insofar as these models inspire and guide their behaviors. These actors – at least in the case of researchers, analysts and regulators – were not those who granted credit to poor American households when everything pointed to the fact that these households would be unable to repay the credit, nor did they guarantee derived products based on these unstable loans. However, through their explicit theoretical choices, or, more often, by tacit acceptance, they encouraged the emergence of these practices and contributed to an aggravation of the consequences.

It is thus crucial to carry out an epistemological analysis of the models that we use to understand reality, due to the consequences that they may have when used to support decisions. Moreover, models also influence our judgments and normative appreciations; there is thus an intimate link to moral innovation.

Responsibility cannot, therefore, be solely considered in terms of possible damage due to a new technology or production process. It must also include epistemological reflection concerning the validity of the theoretical filters that these models apply to reality, starting with the very notions of innovation and responsibility. As we shall see in the following sections, not all innovations contribute equally to the improvement of the human condition. Just because something is new does not mean that it constitutes a form of progress. The importance of the idea of responsible innovation may be interpreted on this basis. It constitutes an attempt to provide a more suitable theoretical and normative framework for the regulation and governance of innovative activities than those currently available (notably in the form of cost–benefit analysis or CSR). By challenging the frameworks on which judgments are based, it paves the way for reflection, which is not always seen in the context of CSR, concerning the very relevance of innovations and the ideal of growth, which they are intended to support.

3.4.1. Understandings of responsible innovation

The idea of RRI emerged independently of work on CSR in the early 2010s, following on from work in the fields of philosophy and sociology of science (notably social and technology studies) to promote different forms of technological assessment [OWE 12, PEL 16a].

Several definitions of RRI have been put forward; these are analyzed in detail in [PEL 16a, PEL 16b] and [GIA 16]. Here, we shall simply provide a brief overview of two types of understandings found in the literature. The first is based on an assembly of substantial propositions and procedural conditions, and corresponds to the different pillars promoted by the European Commission.

In developing a framework program for current and future scientific and technological policy (Horizon 2020, or H2020), the EC included a requirement of RRI, expressed in the form of six “keys”:

• – scientific education;
• – open access to scientific knowledge;
• – public (stakeholder) engagement;
• – gender equality;
• – governance;
• – ethics.

Any scientific project seeking funding from the European Commission must demonstrate a commitment to these different aspects of responsibility, by promoting scientific education and open access to knowledge, ensuring stakeholder participation, guaranteeing and promoting gender equality, implementing responsible governance strategies and taking account of the ethical impact of their activities.

A second type of approach to RRI found in academic literature focuses in the dimensions of responsibility [GRU 11, GRU 12, OWE 12, OWE 13a, OWE 13b, STI 13, GUS 13, FIS 13, VON 12, VON 13, LEE 13, GRO 13, EXP 13, SCH 15, ZIE 15]. While the stresses and the number of cited conditions differ, these approaches may be summarized using five conditions. Innovation and research can be considered responsible when they arise from procedures in which information concerning the developed entity and its environmental and societal effects is allowed to circulate freely (transparency), where risks and benefits are anticipated, and where the actors involved in research and innovation demonstrate a certain responsiveness in adapting to a constantly changing environment, with a proven capacity for reflexivity (i.e. perceiving and analyzing the mental frameworks on which their judgments are based). The final condition is that research and innovation processes should include the different actors concerned by the activity – scientists, innovative companies, public authorities, non-profits, citizens, etc. – in a collective deliberation process, permitting co-construction of the desired aims and standards regulating scientific activity and innovation.

Without going back over the analyses presented in [PEL 15, PEL 16a, PEL 16b], note that this approach to RRI forms part of a more procedural understanding than that presented by the European Commission; in this case, responsible approaches are primarily (but not exclusively) characterized by process quality (transparency, anticipation, responsiveness, reflexivity and deliberation).

Essentially, these understandings offer a different perspective on responsibility to that found in CSR, as presented in the previous chapters.

First – as in the case of CSR – RRI is founded on the idea that stakeholders need to be included in the process of developing and evaluating new technologies (both in the definition put forward by the European Commission and in most academic articles). However, the idea of responsible innovation includes all groups involved in the activity, not only businesses. Each actor (non-profits, political decision-makers, civil organizations and companies) has their own responsibility, which may cover a variety of realities and practices, such as the responsibility of non-profits to propose constructive debates, and that of public authorities to create institutions promoting the emergence of ethical practices. This perspective is therefore more systemic than that encountered in CSR, where the company is at the center of relationships with its stakeholders¹⁶. As we have seen, companies may play a key role in the transmission of good practice, notably along supply or production chains. However, responsibility is still understood from the perspective of corporate actors, who may influence or be influenced by others. From the standpoint of responsible innovation, the whole network of actors is important, and all must participate in the creation and normative evaluation of new technologies or production processes.

Second, the notion of RRI involves requirements that are often only implicitly present in the idea of CSR. First off, in this case, anticipation is not only seen in terms of the strategic positioning of a product by the company, but in a much broader manner, in terms of the different actors involved and the benefits or damages that may result from new activities. As we demonstrated in [PEL 16a, PEL 16b], anticipation in RRI promotes moral innovation, as it also concerns the normative frameworks used to evaluate innovations. In other terms, the anticipation process is not simply a projection of a cost–benefit analysis. It also implies consideration of the normative and regulatory frameworks that may govern innovation, alongside reflections as to the form they may take and the limitations they may impose, from an early stage in the innovation process, at a time when decisions can still be (at least partially) reversed.

One example that demonstrates the importance of discussing normative issues is the recent suspension of sales of Google’s mass-market augmented reality glasses (Google Glasses). Several reasons were put forward for the project’s suspension in January 2015, when the technology was due to be released: the high price, impersonal aesthetics, technical shortfalls and, significantly, concerns around privacy protection. The glasses included a microphone, video camera and Internet access, making it possible to film and upload scenes to the Internet without the knowledge or consent of the subjects. Other suggested applications made use of facial recognition technology, offering, for example, the possibility of consulting a person’s social network profiles while talking to them. Google finally shifted its focus onto professional markets (police, surgeons, etc.) with the aim of responding to specific needs, relativizing the “failure” of the mass-market product. This example demonstrates the crucial role of dialog with stakeholders, enabling them to express and explain their views on new technologies, during the product design process. While the company did provide early users of the product with the option to share their impressions and comments via the Google Explorer Edition program, this group was essentially made up of well-to-do and technologically savvy beta-testers, generally less risk averse with regard to the diffusion of private data than the average consumer. A broader consultation process would have highlighted the legitimate fears and hesitations of consumers, potentially leading Google to modify its target market at an earlier stage in the process to focus on certain specific and regulated uses. Better anticipation of consumer reticence might also have led to the development of a regulatory framework to limit the use of the glasses, or to the development of technical solutions to prevent people from being filmed without their knowledge.

Compared to CSR, RRI also stresses the responsiveness of actors as an essential determiner in terms of responsibility. The need for constant adaptation is clearly highlighted in corporate management rhetoric¹⁷, for example in the perspective offered by Sethi [SET 75] and discussed in Chapter 1, where adaptation and anticipation are seen as a crucial part of stakeholder management (see section 1.4.3.2). However, responsible innovation does not simply require adaptation to market constraints, but also to the different and potentially conflicting demands of social actors. The final purpose of responsiveness in this case is different: in RRI, it is not measured through commercial success or profitability, but rather in terms of the potential benefits (in the broadest sense) to society as a whole. Moreover, as in the case of RRI more generally, companies are not the only entities concerned: each actor, from public authorities to funding bodies, non-profits to civil society as a whole, or, in short, the whole innovation “ecosystem”, needs to be able to adapt and respond to new demands.

Finally, one of the most crucial aspects of RRI, relatively new compared to those found in CSR literature, relates to reflexivity. This condition requires social actors to consider the cognitive frameworks through which they identify problems, develop solutions and evaluate all of these stages in normative terms. This individual, organizational and systemic capacity to explicitly identify cognitive frameworks (theories of knowledge and moral theories, alongside certain presumptions that condition our assessments) is a prerequisite for dialog. It allows us to understand the perspectives from which others express their opinions and eventually to modify our own standpoints; these elements are essential to enable discussion between entities with different and opposing interests and judgments [PEL 16a]. This specific dimension of the conditions for dialog is much less visible in CSR literature than it is in RRI.

To conclude, the idea of RRI aims to adapt to certain specific constraints of innovation that are not necessarily taken into account in CSR. However, this theoretical framework is not immune to criticism. Section 3.4.2 is devoted to different areas in which further work is required in order to create a fuller understanding of RRI, one with the capacity to effectively guide different actors toward more responsible practices.

3.4.2. The conflict between innovation and responsibility

As we have seen, the concept of responsible innovation and the recommendations made within this context are intended to advance the understanding of responsibility within the innovation process. However, certain authors have expressed hesitations with regard to its relevance, going so far as to suggest that the concept of responsible innovation itself is inherently contradictory (echoing criticisms of CSR) [PAV 14b, BLO 15]. Our intention in this section is to explore the difficulties encountered in RRI which have given rise to these criticisms.

3.5.1. Co-construction, participation and responsible innovation

The first recent development in innovation that offers interesting parallels to the co-construction condition included in different approaches to RI is that of open innovation practices. In a work which is now seen as an authority on the subject [CHE 03], then in later works [CHE 06], Chesbrough describes a first model of innovation that is characteristic of the development of big businesses in the 20th Century. He describes this model as “closed”, due to the fact that the processes of innovation (idea creation, production, commercialization, distribution, maintenance and finance) are directly and tightly controlled by the company. This model is notably based on a representation of strategic organization in which a company may take one of three approaches in the face of competition: cost reduction, product differentiation or concentration on a specific market segment [POR 80].

Since the late 1990s, this model has been significantly undermined by increasingly rapid development of technologies and the market possibilities they create. The competitive advantages linked to innovation are increasingly short-lived, and innovation needs to be seen as a strategic dimension of company dynamics to enable the creation of new products and the conquest of new market segments [GOU 12]. However, innovation requires access to knowledge, which evolves rapidly, and the creation of new knowledge that companies may use to generate profit [HSI 12]. A single firm only possesses a limited number of resources for creating new sources of knowledge [KAZ 16]. The idea of “open” innovation takes on its full meaning in this context. Chesbrough defines open innovation as:

“A paradigm that assumes that firms can and should use external ideas as well as internal ideas, and internal and external paths to market, as they look to advance their technology” [CHE 06].

This model emphasizes information sharing and collaboration with other actors, in a climate of confidence and of openness to a highly varied “outside”. This “outside” includes other companies [VAN 06], other individuals, including external experts or expert networks [ROH 10], or even communities of practice [GOU 12]. Furthermore, different types of individuals may become involved in the innovation process: consumers [HOY 10, MAR 14], suppliers, competitors, academics, public-sector actors, members of non-profits or other interest groups; each person brings their own complementary knowledge and experience to the table, in the aim of co-creating value [RAM 10, GOU 12, BOG 11, DRI 13]. Innovation is perceived as a “multiplayer game”, bringing together a network of actors in a complex system of interactions between these actors and their institutional, political and social environment. The central aspect here is not so much the creation of knowledge as its circulation [BES 13].

However, open innovation practices are not free from difficulties; certain authors have even identified an open innovation paradox [BOG 11, GOU 12], similar to the information paradox described by Arrow [ARR 62]. The benefits of open innovation (considerably increased access to knowledge and to rapid developments in a domain) must be weighed against the risk of divulging strategic information to competitors, undermining the competitive advantages that might have been acquired through the collaborative innovation process [GOU 12, SAL 14]. More precisely, as open innovation is based on collaboration with other parties, it requires careful management of which elements may be divulged and which must remain confidential. When any party is excessively reticent in information exchanges, potential partners may fail to perceive any benefits to be gained from collaboration and lose any desire to participate in the process. Exchanging too much information, on the other hand, can leave companies too exposed to competition, and there is a risk that ideas might be used by others before the originating company has had the chance to benefit from them. Identifying an appropriate level of openness with regard to intellectual property is one of the major challenges involved in open approaches [CHE 03, BES 13, MAL 11]. Furthermore, while it represents a step toward co-construction of new technologies, open innovation also creates a whole host of organizational difficulties [ALE 13, FOS 11]. When tasks and boundaries within a company are destabilized, it becomes difficult to coordinate large numbers of more or less active contributors, and to ensure that knowledge is shared and correctly attributed within the group.

From an individual perspective, Salter et al. [SAL 14] have identified several problems facing company executives who wish to open their R&D processes to external collaboration. First, as we have seen, the management of information sharing and intellectual property may cause potential partners to hesitate; businesses are often reticent in entering into mutual engagements without a confidentiality agreement. Creating an agreement of this type can be a long, laborious and costly process (involving extensive discussions between company legal departments, for example), with no immediate guarantee that it will benefit the company. Second, individuals may find it difficult and costly to engage in new forms of collaboration in cases where other, tried and tested, options are already available. Finally, when companies bring in external actors to solve specific technical problems, simply acquiring new knowledge is rarely enough; significant effort is required to translate this knowledge into a form that can be understood by the company, i.e. which conforms to preexisting thought categories [SAL 14].

One way of circumventing the information sharing problem involves using different forms of collaboration to test the potential of a partnership before committing to a more binding relationship. Salter et al. [SAL 14, p. 85] speak of the idea of transaction-light partnerships: before entering into large-scale collaborative activities with external actors, potential partners begin by working together on small, interesting but not necessarily essential projects which, notably, do not require any specific agreements to be made in terms of intellectual property. This allows potential partners to evaluate their capacity to collaborate to their mutual advantage.

Taking a different approach, Gould [GOU 12] draws on literature on stakeholder engagement, which, as we have seen, also considers the ways in which collaboration with actors from outside a company may be a source of value creation [FRE 10]. By fostering dialog with stakeholders, the company acquires knowledge of these stakeholders [SHA 05], enabling better identification of possible risks and the development of relationships of trust with their partners. They also gain technological knowledge, developing new networks of actors able to assist in solving certain problems, and remaining abreast of new, potentially useful, technological solutions [REY 16]. In this way, the company learns to manage knowledge flows (input, output, learning and disclosure) through dialog [GOU 12, AYU 06] in order to create situations that benefit both parties [PLA 09]. Saguy et al. cite an example of successful open innovation led by the Syngenta food company, which developed partnership management mechanisms based notably on financing academic innovation centers. These centers provided a new means of acquiring external technological capacities, promoting long-term relationships to reduce the problems of intellectual property sharing, as discussed earlier [SAG 16, p. 5; MAL 11].

The difficulties raised by appropriate information sharing within the context of open innovation can therefore be overcome using a range of practical solutions, which should be considered on a case-by-case, contextual basis. With regard to the criticisms mentioned in section 3.4.1, the requirement for transparency involved in responsible innovation and the innovation process itself are not radically incompatible. However, conceptual and practical efforts are required in order to develop forms of information disclosure that respond to a number of contradictory criteria.

Another widely studied form of corporate “opening”, running parallel to open innovation, is based on the inclusion of users from a very early stage in the technological design process (user led innovation) [MAR 14]. A variety of tools, such as codesign and living labs, allow potential users to test an innovation and influence its development. Users thus become cocreators of a product, as in the case of Adidas, which allows customers to design their own shoes by combining several different elements (color, print, shape, etc.); the company then produces and delivers the product. Taking a similar approach, Daimler-Benz launched a competition to design its Smart model, attracting responses from all over the world. Participants were able to promote their own design and vote for models proposed by others [BES 13]. This example of crowdsourcing (where companies harvest external input via an online platform) allowed the company to collect new ideas, but also to create a community of active users that could be remobilized at a later date for other forms of participation.

The popularity of living labs is another interesting aspect of open innovation. Developed at MIT in the 1990s by William Mitchell, followed by Kent Larson and Alex Pentland, they allow users to test products in situ, in the context of their everyday lives. This forms a sharp contrast with the traditional laboratory or test panel methods used by companies. In the healthcare sector, for example, the Sainte Justine pediatric university hospital in Montréal worked with the Société des Arts Technologiques (SAT) to create a living lab in order to test new therapeutic tools, designed in association with users and in response to their specific needs. One result of this was the Marionnect project, in which children with communication problems are able to interact with a life-sized avatar of their own choosing, rather than with a human psychiatrist, as an intermediary step toward more traditional face-to-face methods of communication. In this case, the technology was developed in direct collaboration with the different actors targeted by the tool, and tested and modeled in response to the requirements and assessments expressed by these users.

These different examples of new models of innovation breathe life into the theoretical requirements for technological co-construction and stakeholder participation formulated by proponents of RI²¹. They also highlight the benefits of this approach, both for users, who are able to mold products according to their own preferences, and for companies, in the form of guaranteed market opportunities.

However, consulting and involving users, and partnerships with other stakeholders more generally, do not guarantee effective engagement of actor responsibility. These examples of open innovation involve two types of responsibility: legal responsibility, in connection with the terms of a research partnership or a confidentiality agreement, for example, and responsibility in terms of responsiveness, which is notably engaged in cases where actors find creative solutions to the information paradox. Careful dosage of information disclosure in order to encourage collaboration without fundamentally undermining commercial advantages is, in itself, an art form that requires skills and experience. However, in these different examples, the questions of possible damage and the normative assessment of the resulting product are not considered. Co-construction (of a technology, for example) and stakeholder participation are lauded for the beneficial effects that they may have in terms of value creation; however, nothing has been said regarding (1) levels of actor awareness of the social and environmental impact of the technologies in question, or (2) the underlying worldviews that justify the interest of the product. There is therefore a tension – analyzed in detail in [PEL 16b] – between procedural and substantive responsibility, clearly visible here. Ensuring that the correct “ingredients” of a production process are present is not always enough to make it responsible; in addition to the conditions of responsibility, a consequentialist evaluation is sometimes required, in connection with the attainment of certain goals [PEL 16b]. As we stated in our work with Bernard Reber [PEL 16a], moral pluralism – understood, in this case, as the mobilization of a variety of normative theories – is therefore essential in effectively fostering greater responsibility in innovation practices.

That said, innovation practices that are open to participation from final users, living labs and other new forms of corporate collaboration provide valuable opportunities for open discussion and normative assessment of technologies, and of the innovations in question more generally. However, for responsibility to be present, the normative assessment element must feature explicitly in the specification, something that is not necessarily true in the cases mentioned above. This conclusion reflects that of the previous section: the idea of normative discussion of the results of collaboration must be integrated into the process alongside dialog, openness and co-construction. This implies moral assessment of these results in the light of the value systems and normative theories applied by different actors. The idea of responsible innovation is essential in this case: combined with existing practices, it adds a necessary ethical dimension.

3.5.2. “Social” innovation

The idea of “social” innovation is a final development in innovation practices, which, by its very nature, clearly reflects the requirements of responsible innovation.

The idea appears to have originated in the 1960s [MOU 13, p. 13], although the term itself did not emerge until the 1970s [CLO 03]. The notion of SI is even more polysemic than that of innovation alone, covering a range of potentially different efforts and requirements. The addition of the “social” element makes the term even harder to qualify.

First, in a broad sense, the term “social” denotes “that which relates to society”, as in the case of the “social” sciences. In this case, SI corresponds to the organizational transformations of a company or organization [HIL 04] or to developments in civil society, social movements or the State [HOC 09]. However, the distinction between innovation and SI in this case is too weak [GRO 14]; according to this interpretation, almost any innovation might be considered “social”. SI is more usually defined according to its purpose: the “social” qualifier relates to an improvement in the quality of working conditions, improved management of natural resources, or any other evolution considered to be socially desirable.

For Moulaert et al., for example, SI involves:

“finding acceptable progressive solutions for a whole range of problems of exclusion, deprivation, alienation, lack of well-being, and also to those actions that contribute positively to significant human progress and development. SI means fostering inclusion and wellbeing through improving social relations and empowerment processes: imagining and pursuing a world, a nation, a region, a locality, a community that would grant universal rights and be more socially inclusive” [MOU 13, p. 16].

It thus involves finding “solutions” to the specific problems posed by certain vulnerabilities (poverty, disability, isolation, age, etc.) that are acceptable to a large group of individuals, i.e. which are not only effective in economic terms. Innovation may relate to a new product, a new manufacturing process that reduces product cost and makes it accessible to new sectors of the population, or to new organizational or marketing procedures. Examples of SI include microfinance, generic drugs and low-cost ICT solutions, accessible to the poorest members of society. SI also often includes a local dimension: innovations are linked to a specific territory, at community level, within a specific context [DJE 12], and include the actors they are intended to benefit, as in the case of the Danone–Grameen partnership discussed in Chapter 2.

Generally speaking, “social” innovation thus involves a substantive form of responsibility (see Chapter 1). It relates to a moral purpose based on a particular understanding of the “good” with definable contents: improving the living conditions of workers or certain categories of vulnerable individuals, shifting focus away from the unrelenting pursuit of profit, etc. While the intentions behind SI are perfectly laudable, this type of approach raises difficulties in terms of determining criteria of the good (which may be a source of conflict) and, in a related manner, to paternalistic attitudes to the definition of the good.

The example of “golden rice”, created at the Swiss Federal Institute of Technology in Zurich in the late 1990s, is a case in point. Produced using a series of genetic modifications, golden rice²² is enriched in β-carotene (a precursor of Vitamin A) and is aimed at reducing malnutrition linked to a shortage of dietary Vitamin A in countries where rice makes up the staple diet. Its proponents considered it to be an essential SI for combatting the effects of Vitamin A deficiency in Asia (notably blindness and early death). Its critics feared the generalization of GM foodstuffs. The Indian activist Vandana Shiva, for example, denounced the methods used to produce this GM rice (monoculture, environmentally damaging herbicide and pesticide use, costly patents, etc.), while other critics focused on the risks of excessive dependence on a single foodstuff²³. While some of these criticisms were somewhat extreme and were accused of closing dialog, as in the case of the Greenpeace campaign against the project²⁴, which denounced all approaches based on genetic modification, they did raise an essential point: science and innovation alone cannot provide a full response to a complex phenomenon like malnutrition, which also has political and institutional roots. Promoting an innovation such as golden rice without considering institutions, the reasons poor individuals are unable to access basic resources (many people are undernourished in India and Thailand, for example, despite the fact that these countries are some of the biggest exporters of rice), and the impact of this type of crop on the country’s agriculture as a whole does not constitute a fully responsible strategy. The presentation golden rice as the “magic pill” to solve the problem of malnutrition in Asia demonstrated a techno-centric, conquest-based attitude, which understandably irritated certain stakeholders (non-profits and civil society). The conditions of reflexivity and inclusion, essential to RI (section 3.4), are only partially satisfied in cases where certain parties dismiss the legitimate concerns of others from the outset.

Finally, the fact that innovation has a social bent does not necessarily make it beneficial for those for whom it is created. Once again, the means of producing and commercializing an innovation that appears promising, but not risk-free, can only be defined by pluralist normative assessment. Dialog and pluralist reflection, including at least some of the actors affected, are necessary in order to identify that concerns expressed by hesitating parties are legitimate criticisms, which are strategic and which are simply opportunistic demands, and to create technological, but also social and political responses to the problems which initially motivated the innovation. SI is a strong motivating force for responsible innovation, but it does not make use of all possibilities due to its foundation on substantive and debatable understandings of the good. Without a certain level of co-construction and stakeholder engagement, even the best intentions and the most enthusiastic support can result in authoritarian and blind dynamics, compromising the initial good intentions. Ends and means must therefore be considered together in order to guarantee responsibility in innovation.