Secular AI ethics: Building a universal code

We live in a world where AI continues to influence various facets of our lives. Given this context, the ethical considerations surrounding AI’s development, deployment, and use are growing increasingly intricate. With the vast diversity of cultural, religious, and philosophical beliefs worldwide, creating a universally accepted ethical framework for AI is challenging. This is where secular AI ethics become essential. It’s an approach rooted in shared human values transcending specific religious or cultural doctrines.

What’s secular ethics?

Secular ethics refers to a system of principles independent of religious beliefs. Instead, it focuses on universally applicable values derived from reason, experience, and shared human concerns. Secular ethics emphasises the importance of rational thought, human welfare, and the collective good rather than relying on divine authority or religious texts to dictate right and wrong.

Secular ethics is not exclusive to any particular group or belief system. It’s a system that values justice, fairness, equality, and human rights, making ethical decision-making inclusive and accessible to people of all beliefs. This inclusivity ensures that these principles have universal applicability and can be understood and respected by all.

Key philosophers and their contributions

Traditionally, commentators trace the foundation of secular ethics to the Age of Enlightenment. According to them, during the Enlightenment period (in the 17th and 18th centuries), thinkers advocated for reason, science, and humanism as the basis for ethics rather than religious dogma.

Some key philosophers who have shaped secular ethics include the following:

Immanuel Kant (1724–1804)

Kant’s philosophy centred on the categorical imperative, which argues that moral principles should be universal and applicable to all rational beings. He believed that actions are morally right if they can be universally applied and respect the inherent dignity of individuals.

John Stuart Mill (1806–1873)

Mill was a proponent of utilitarianism, a philosophy suggesting the best ethical action is the one that maximises overall happiness or minimises suffering. His work emphasised the importance of consequences and the collective well-being of society.

John Rawls (1921–2002)

Rawls introduced the concept of justice as fairness. This concept argues for an ethical system based on equality and impartiality. He proposed the ideas of the original position and the veil of ignorance as tools for determining the fairness of ethical principles. Let’s illustrate these ideas.

Imagine you’re in a group deciding on the rules for a new society.

• Original position: A hypothetical situation where everyone is equal, and no one knows their own future position in society. You don’t know if you’ll be rich or poor, healthy or sick, powerful or powerless.
• Veil of ignorance: The mental barrier preventing you from knowing anything about your circumstances in that new society.

The idea is that if you’re making decisions from behind this veil, you’ll be more likely to choose fair and just rules. Why? Your own personal situation won’t bias you.

Peter Singer (1946–present)

Singer is a contemporary philosopher known for his work on effective altruism and ethics concerning animals and global poverty. He advocates considering the interests of all beings capable of suffering, extending moral consideration beyond just humans.

Re-contextualising the Enlightenment

It’s worth noting that while these philosophers have made significant contributions, the Enlightenment itself was a complex period with both positive and negative impacts on the world. It’s also crucial to recognise that expressions of reason, humanism, and ethical thought existed in various cultures long before and alongside the European Enlightenment. For example, consider the following:

• Golden Age of Islam: Spanning roughly from the 8th to the 13th centuries, this period saw significant advancements in science, mathematics, philosophy, and literature across the Islamic world. Thinkers like Ibn Sina (Avicenna) and Ibn Rushd (Averroes) engaged in deep philosophical inquiry, exploring topics like reason, ethics, and the relationship between faith and knowledge. Their works influenced both Islamic and European intellectual traditions.
• Ancient Greece: The philosophers of ancient Greece, such as Socrates, Plato, and Aristotle, laid the foundations for Western philosophy and explored questions of ethics, justice, and the nature of reality. Their ideas continue to resonate today and have influenced countless thinkers throughout history.
• Confucianism and Daoism: These ancient Chinese philosophical traditions emphasised values like humaneness, righteousness, propriety, wisdom, and the pursuit of harmony with nature. They provided ethical frameworks for individuals and societies, influencing governance and social relations across East Asia for centuries.
• African Philosophical Traditions: Pre-colonial Africa had rich and diverse philosophical traditions that explored ethical concepts, social organisation, and the nature of reality. Although often oral in transmission, these traditions laid the groundwork for later African intellectual movements and continue to inform contemporary African thought.

Further, post-colonial theory reminds us that the legacy of colonialism continues to shape power dynamics and knowledge production. In the context of AI ethics, this means actively challenging Eurocentric biases and ensuring that diverse voices and perspectives are included in the conversation.

It’s my view that while each philosopher has their contributions and criticisms, the point is to consider the utility of their ideas in shaping our understanding of ethical AI development.

Examples of secular ethics codes

Secular ethics has informed the development of various ethical codes and frameworks that guide decision-making in different fields.

Some notable examples include the following:

Western

1. The Universal Declaration of Human Rights (UDHR): The United Nations adopted the UDHR in 1948. It articulates universally protected fundamental human rights based on equality, dignity, and respect for all people, regardless of their religious beliefs.
2. The Belmont Report (1979): This report outlines ethical principles and guidelines for research involving human subjects in the US. It’s grounded in respect for persons, beneficence, and justice. It also aims to protect individual rights and well-being in scientific research.
3. The European Convention on Human Rights (ECHR): Established by the Council of Europe in 1950, the ECHR is an international treaty to protect human rights and political freedoms in Europe. The convention is based on democracy, the rule of law, and the protection of individual rights.
4. The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems: This initiative provides guidelines for the ethical design and implementation of AI and autonomous systems. It emphasises values like transparency, accountability, and fairness, aiming to ensure that AI technologies benefit humanity.

Eastern

• The Tirukkural: This ancient Indian text embodies ethical principles that transcend religious boundaries, making it a valuable contribution to Eastern secular thought.
• The Constitution of India: India’s constitution, with its secular framework and emphasis on individual rights and equality, represents a significant Eastern example of secular governance principles.

African

• Ubuntu philosophy: This indigenous African philosophy, with its focus on interconnectedness and community, offers a unique perspective on secular ethics rooted in African values.
• The African Charter on Human and Peoples’ Rights: This charter, emphasising collective well-being and solidarity alongside individual rights, represents a Pan-African approach to secular ethics.

Secular ethics in AI

Applying secular ethics to AI development involves grounding AI systems in these shared human values like:

• respecting human rights,
• promoting fairness, and
• transparency and accountability.

Practical implementation of secular AI ethics

Practising secular AI ethics requires a multi-pronged approach, encompassing the entire AI lifecycle.

1. During the design phase, developers should integrate ethical considerations into the core algorithms. This approach ensures that AI systems don’t perpetuate biases or discriminate against specific groups.
2. All role players in the AI lifecycle should prioritise transparency and explainability. This would ensure that users understand how AI systems reach their conclusions.
3. Regular audits and impact assessments can help identify and rectify potential ethical concerns.
4. Promoting diversity and inclusivity within the AI development community encourages diverse perspectives and avoids unintended biases.
5. Creating accessible channels for feedback and redress empowers users to report any ethical concerns they encounter, fostering a culture of accountability and continuous improvement in the AI landscape.

Contextualising secular AI ethics to local needs

As shown above, secular AI ethics provides a universal foundation making decisions about AI. Still, contextualising ethics to address individual countries’ unique needs and challenges is crucial. Each nation has its distinct social, economic, and cultural landscape. And we must consider this landscape when applying ethical standards to AI.

1. Addressing local socio-economic realities: For example, in a country with significant economic inequality, a secular AI ethics framework might emphasise fairness and access more strongly. The intention is to prevent AI technologies from widening the gap between different socio-economic groups. This process could involve specific guidelines to prevent the misuse of AI in ways that could exacerbate existing disparities.
2. Respecting cultural sensitivities: In countries where cultural practices and traditions play a vital role in society, we must develop AI systems to respect these cultural nuances and adhere to universal ethical principles. This means engaging with local communities and stakeholders to ensure that AI deployment aligns with both global ethics and local values.
3. Navigating legal and regulatory landscapes: Each country has its own legal framework, and secular AI ethics must be adaptable to fit within these regulatory environments. This might involve creating country-specific codes of conduct that incorporate both international ethical standards and local laws, ensuring compliance and ethical integrity.
4. Promoting ethical AI in developing economies: In developing countries, where technological infrastructure and digital literacy might still be emerging, a secular AI ethics framework could prioritise education and awareness. This approach could ensure that countries introduce AI to empower rather than exploits local populations. This includes fostering transparency in AI technologies to build trust among users unfamiliar with advanced digital systems.

Moving forward

Developing secular AI ethics is not about creating a one-size-fits-all solution. Instead, it’s about crafting flexible guidelines to accommodate different perspectives while upholding shared values. As AI continues to evolve, so must our ethical considerations. Society must ensure the technology serves humanity in a fair, just, and inclusive way.

At ITLawCo, we are committed to contributing to the dialogue around AI ethics. We want to work towards frameworks that respect diversity, promote transparency, and ensure accountability. As we move towards an AI-centric future, establishing a secular ethical foundation will be crucial in navigating the complexities of this technology.