Oluwadamilare (Eyinimofeoluwa) Odusami

Ethics and the Evolving AI Landscape: The morals of AI Systems

 Ethics Simply means the ability to determine right and wrong, it is the standard guiding our behaviors based on moral principles. The term Ethics has been very controversial in the Information technology sector, but ethics aims to create a strong sense of principles that govern behavior or conduct. In simple terms, ethics determines the best ways we interact and live as humans while adopting technological advancements to solve human problems.

As cybersecurity professional we would most often get privileges and roles that will require access to sensitive information and to this note, it is important that as security professionals, we reflect on the ethical implication of the decision we or the systems we create take.

Cybersecurity ethics refers to the moral principle and standards that govern the conduct of cybersecurity professionals and the best practices that aims at safeguarding data, computer systems and networks from unauthorized access and attacks.

The major aim of any ethical practice is to uphold the rights and interest of individuals, organizations and societies while also adhering to the legal and regulatory frameworks, governing data protection and cybersecurity.

Ethics are paramount because they help preserve the integrity, confidentiality, functionality and reliability of human institutions and service that depend on data, Information systems and networks. By upholding ethical principles, cybersecurity professionals can mitigate risk and promote proper use and governance of Information systems, which will in turn create a safer and more resilient digital environment for all stakeholders.

Artificial Intelligence is the ability of a system to perform cognitive functions associated with human minds. Such cognitive functions may be reasoning, learning and problem solving. There are basically two types of AI systems;

• Rule based AI (Expert AI), which behaves according to a set of fully defined rules that has been stipulated by the trainer.
• Learning-based AI, which solves problems and adapts its functionality on its own, based on configurations and datasets.

AI models are mathematical frameworks created by people and trained on data that enables AI systems to perform certain tasks by identifying patterns, making decisions and prediction of outcomes.

AI systems represents a complex structure of algorithms and models designed to mimic human reasoning and deliver services autonomously.

AI Ethics are a set of values and principles that uses widely accepted standards of wright and wrong to guide the moral conduct in the development, deployment, usage and sales of AI system or solutions.

AI ethics strives to study how to optimize and fully utilize AI’s beneficial impacts while reducing the risk and adverse outcomes it poses to humans and services. It is the set of principles that govern AI’s behavior in terms of human values. AI ethics helps to ensure that Artificial Intelligence is used in ways that would positively influence and augment human intelligence and not replace them. AI ethics helps us to ensure there are no barriers and creativity is encouraged while developing our various artificial intelligent system.

While we design and innovate next generation systems, it is important to note that, for us to embed AI technologies in our processes, we have to be mindful of the biases that sweep decisions making. These biases may vary from sector or services and may allow our systems to make decision that would not be fair or efficient.

To control this biases, we have to ensure 3 major principles;

• Principle of Accountability: This principle states that someone should be responsible for the outcomes of AI systems. Accountability denotes that there should be a form of responsibility for the outcomes and impact of AI systems. The principle ensures trust and mitigates the potential harm.
• Principle of Fairness: This principle ensures that AI systems treat people or services equally, and avoids any form of discrimination. Fairness ensures that no group is favored over another especially in terms of race gender, or other social economic status.
• Principle of Transparency: This asserts that the decisions made by AI systems should be explainable and comprehensible. This principle should be applied when selecting algorithms and validating processes. Transparency touches all the processes in the development of an AI system (i.e AI lifecycle). It is important AI systems ensure fairness to help tackle bias in the processing of data.

Although there is no single law or standard that is universally agreed upon to lead the pace in setting ethical AI principles. Many organizations and government agencies consults with professional in the domain to create their guiding principles and frameworks. In developing this principles, we may have to address some of the following issues;

• Human wellbeing and dignity: The systems should always prioritize and ensure the wellbeing, safety and dignity of individuals. The systems should not replace humans nor compromise the welfare of the humans.
• Human Oversight: AI needs human monitoring at every stage in development and adoption. This is because the major ethical responsibilities lie with the humans that runs this system.
• Bias and Discrimination: This can be prevented by ensuring diverse data sets and methods.
• Transparency and explainability: To mitigate the issues in Transparency, we should create systems that users can easily understand and verify.
• Data privacy and protection: We should ensure data in AI systems are securely managed and complies with the various privacy laws. The systems should use encryption, and secure protocols to safeguard data integrity and confidentiality.
• Promoting inclusivity and diversity: AI technologies need to reflect and respect the vast range of human experiences and identities.
• Society and Economies: Artificial Intelligence should help drive societal development and economic advancements for everyone, without fostering inequality or unfair practices.
• Enhancing digital skills and literacy: AI technologies should strive to be accessible and understandable to everyone, regardless of the person’s digital skills.
• Business Welfare: AI systems should prioritize the processes that aid the day to day running of businesses and also maximize efficiency, while creating room for scalability.

Responsibilities in AI Ethics

In developing Artificially intelligent systems. There lies an ethical question that should be known to all “Who is responsible for AI ethics?”. A simple answer to this question is “Every stakeholder involved in the development and use of the system”. These include governments, consumers, citizens and even business that integrates them into their system.

Below are Some of the Different Roles of People in AI ethics.

Academic Institutions: Their major role is in research and educational advancements in the development of ethical guidelines that surrounds AI systems.

Policymakers and Regulators: These are mostly governmental institution or independent bodies that create laws and regulations to govern the ethical us o AI and protect individual rights.

Businesses and Industry Leaders: ensure their organizations adopt ethical AI principles so that they are using AI in ways that contribute positively to society.

Developers and Researchers: They play a crucial role in creating AI systems that respect and lay emphasis to human agencies and oversight, they also address bias and discrimination, and are transparent and explainable

Civil Society Organizations: Their major role is to advocate the ethical use of AI, and play a role in oversight, and provide support for affected communities.

End Users and Affected Systems: These people play a role in ensuring that the Ai systems are explainable, interpretable, fair, transparent and beneficial to society.

Best practices for forming an AI ethics steering Committee

• Composition and Expertise: The Ai ethics steering committee should include members form diverse backgrounds, which may range from geographical location, ethnicity, expertise and gender. This very important because Ai systems influences the lives of people in diverse background and spheres of life.
• Defining the Purpose and Scope: The purpose and scope of an AI system should establish a framework for creating a responsible governance structure in the development and deployment of AI technologies.
• Defining Roles and Responsibilities: When roles and responsibilities are clearly defined there would be effective distribution of responsibilities, making accountability easier.
• Setting Objectives: This would ensure alignment with the values and success metrics of an organization. Organizations can also set up frameworks for creating objectives that will aid development and other ethical principles. It is also important that these objectives are specific, measurable, achievable, relevant and time-bound.
• Creating Procedures: This procedure will help AI steering committees to operate in a more consistent and effective manner. The procedures should be clear, accessible, flexible, and accountable.
• Ongoing Education and Adaptation: This is critical in maintaining relevance and effectiveness of in an AI steering committee. It will allow the committee to stay informed about the evolving landscape, regulations and ethical principles while also striving to have a structure for improvement.

Key Technical Requirements for ethical AI systems

• The systems should be able to detect bias and be explainable.
• Ethical AI systems should ensure privacy and security of data.
• The systems should undergo continuous monitoring and updating.
• The systems should adopt stakeholder engagement and feedbacks.

Key Components of an effective Ethical AI training curriculum

• Provide Foundational Knowledge of Ethical AI and AI systems
• Comprehensive curriculum development: The should be a provision of well-rounded understanding of AI concepts and application, regardless of the role they play in the development of the AI system.
• Role specific training modules: The training curriculum should be tailored to the needs of the specific roles, which will ensure the participants develop skills that will be applicable to the role they perform.
• Interactive learning experiences: The training curriculum should show case real life scenarios and should practical insights in the development of Artificial intelligent systems
• Assessment and Certification: the learning outcomes should be able to be measure and certifications should be provided in order to recognize the achievement of these outcomes.
• Feedback Mechanisms: Effective AI training curriculum should be tailored to the audience’s needs, irrespective of if they are beginners, intermediate learners or advanced practitioners.