In the contemporary corporate scenario, deceptive practices have been spreading, showing themselves under a spectrum of several “colors,” each accounting for a specific form of corporate, ethical, and social dissimulation. This phenomenon, so-called washing, reflects a worrying trend of organizations that seek to project an image of responsibility and awareness without actually implementing substantial changes in their operations and core corporate values.
Perhaps the most widespread of these trends is greenwashing, which could be defined as adopting an environmentalist rhetoric without the expected commitment to genuinely sustainable practices. Companies engaged in this practice often invest more funds in environmental marketing, and in building a green image, than in concrete efforts to mitigate ecological impact, creating an illusion of responsibility, while merely boiling down to market engagement strategy.
Likewise, such scenarios also affect gender and diversity issues with pinkwashing and purplewashing, i.e., addressing the instrumentalization of important social causes for corporate branding purposes only. Pinkwashing (or rainbow-washing) manifests itself through the apparent defense of LGBTQ+ rights, often limited to specific periods such as Pride month, without a substantive commitment to inclusive policies throughout the year, and systematically in a given business’ corporate programs. Purplewashing, in turn, uses the flag of gender diversity and inclusion in a shallow way, without implementing structural changes in hiring, promotion and organizational culture policies.
The color palette of washing practices also extends to bluewashing, which accounts for a form of manipulation focused on digital and corporate ethics. Organizations that resort to this strategy often claim adherence to strict ethical principles in the use of technologies and data, while in practice they may be involved in questionable practices of collecting and using personal information, or even the lack of a de facto corporate policy on data protection and technology management, as in the case of artificial intelligence.
Such practices, taken together, not only damage the relationship with the audience and stakeholders, but also erode the credibility of bona fide corporate responsibility initiatives.
In this context, and with the implementation of layers of artificial intelligence at the most diverse management levels and in the most diverse business areas, it is of paramount importance to define corporate policies that are properly attached to the legal dictates and values of the company. In addition, such policies should provide for effective instruments and methods of implementation in order to prevent them from becoming another pillar of bluewashing movements.
Bill 2338/2023 in Brazil represents a significant milestone in the regulation of artificial intelligence, in addition to the fundamental provisions (rights and duties of those involved, for example), the Bill aims to establish key guidelines for the implementation of effective governance programs. Like European regulation, called the AI Act, the bill emerges in a context of growing global concern about the ethical and social impacts of artificial intelligence, recognizing the urgent need for regulatory frameworks that can keep up with the rapid pace of technological development.
As for compliance and corporate ethics programs, the bill requires organizations that develop or implement artificial intelligence systems, referred to as artificial intelligence agents, to establish comprehensive governance programs to ensure transparency, explainability, protect the fundamental rights of users, mitigate risks associated with the use of such systems, and further a culture of responsibility and accountability.
Transparency and explainability, cornerstones of these governance programs, imply the clear and understandable disclosure of the details of the operation of artificial intelligence systems, as well as the criteria used and the logic adjacent to their decision-making processes. This openness not only allows for more effective public scrutiny, but also boosts users’ trust in the implemented systems.
Another crucial aspect addressed by the legislation is the protection of fundamental rights, as it requires organizations to implement robust safeguards against potential privacy violations, algorithmic discrimination, and other negative impacts that artificial intelligence systems may have on individual and collective rights. This entails conducting regular algorithmic impact assessments, and implementing human oversight mechanisms that can intervene in cases of potentially harmful decisions.
At this point, a particularly relevant point of the bill is the emphasis on the need for corrective measures to avoid the perpetuation and expansion of structural social biases by technology. Recognizing that diversity in the development team can significantly contribute to reducing algorithmic biases, the bill innovatively proposes the requirement of an inclusive and diverse team responsible for the design and development of artificial intelligence systems. This approach not only fosters equity in the process of technological creation, but also serves as a preventive measure against the inadvertent incorporation of biases into algorithms.
Bill 2338/2023 also points up the need for continuous training of teams engaged in developing and operating artificial intelligence systems. This requirement recognizes that effective governance of artificial intelligence is not a monolith, but a dynamic process that requires ongoing updating of knowledge and practices. The bill also provides for the creation of codes of good practices and governance by artificial intelligence agents, covering aspects such as organization, operation, security, and risk mitigation. This approach indicates a commitment to responsible self-regulation and adaptability to the specifics of different artificial intelligence applications.
True ethical governance in artificial intelligence cannot be reduced to a set of procedures or compliance checklist. It demands a profound cultural transformation, where ethical principles and respect for fundamental rights become intrinsic to the organizational ethos . This implies the need to foster an environment where ethical issues are constantly debated and considered at all decision-making levels, from the initial design of an artificial intelligence system to its implementation and continuous monitoring.
The development and effective application of governance programs in artificial intelligence, as advocated by the Bill, represents a significant challenge for organizations, requiring an approach that transcends mere legal compliance and is anchored in a deep and genuine ethical commitment. To prevent these governance efforts from sinking into another manifestation of bluewashing, specifically in the context of artificial intelligence, it is imperative that organizations adopt a proactive and committed approach to ethics and responsibility in the development and application of these technologies.
In this context, the challenge posed by Bill 2338/2023 should not be seen merely as a regulatory obstacle, but as an opportunity for organizations to reaffirm their commitment to social and ethical responsibility. By fully embracing the principles of ethical governance in artificial intelligence, companies have the chance to paint an authentic picture of integrity and compliance, distancing themselves from deceptive washing shades. Thus, instead of masking its practices with superficial colors of compliance, this new artificial intelligence governance palette made up of genuine displays of transparency, accountability, and respect for fundamental rights, not only aligns organizations with legal and social expectations, but also causes companies to create a vibrant spectrum of ethical innovation, paving the way for a technological future that reflects, in its deepest nuances, the values of a truly equitable and beneficial society for all.
Autor:
