On October 21, 2024, I had the pleasure of delivering the evening keynote address at the 19th International Conference on Knowledge Management (ICKM), titled “From Data to Justice – Creating Ethical AI Through Knowledge Management”. The following details and expands on what was discussed during my keynote address.

Today, data is the new currency that drives decision-making across sectors—from healthcare and education to criminal justice and finance. AI, powered by vast datasets, holds tremendous promise. However, that promise comes with significant responsibilities. The data we use, the models we build, and the algorithms we design must operate ethically and equitably. Bias in AI systems, whether intentional or inadvertent, can lead to discrimination and social injustice. To bridge the gap between data and justice, we must foster an environment where ethical AI is not just a concept, but a reality.

This is where knowledge management (KM) becomes critical. Effective KM ensures that the right data is collected, verified, and contextualized before being fed into AI systems. By establishing transparent processes for data governance, we can mitigate biases and enhance accountability. Knowledge professionals must work alongside data scientists to curate diverse, representative datasets that reflect the complexities of society.

Knowledge Management as a Pillar of Ethical AI requires more than just technical solutions—it demands a multidisciplinary approach that includes subject matter experts, diverse business stakeholders, business analyst, sociologists, law, and knowledge management. KM serves as the backbone for documenting ethical standards, best practices, and governance frameworks within AI development. Through robust KM practices, organizations can institutionalize ethical guidelines that govern AI lifecycle stages, from data collection to algorithmic deployment.

Knowledge Management, in its essence, is about harnessing the collective know-how of organizations, aligning processes, and ensuring that valuable information and knowledge is accessible, usable, and ethical. By embedding KM in the development of AI systems, we create a foundation that prioritizes transparency, accountability, and integrity.

Imagine a scenario where AI developers have access not only to high-quality datasets but also to curated ethical guidelines, documented lessons learned from previous projects, and a diverse set of perspectives on the potential impacts of AI. Knowledge management ensures that these critical pieces of information are woven into the AI lifecycle—from data collection, model selection, model development, model training, and model deployment.

To create truly ethical AI, we must take a holistic approach that integrates knowledge management into every phase of the AI journey. It starts by incorporating diverse perspectives that are reflected in the AI stakeholders and teams. This leads to determining the use cases to apply AI, model selection and then to the critical process of data selection, during which we understand the data we collect: its sources, its potential biases, and its limitations.

By managing our knowledge effectively, we can identify gaps, ask the right questions, and mitigate risks before they manifest in AI systems that affect people’s lives. Furthermore, knowledge management enables collaboration across disciplines—bringing together ethicists, domain experts, technologists, and policy makers—to co-create AI solutions that are just and equitable. It ensures that ethical considerations are not an afterthought but are embedded into the very DNA of the technology.

Knowledge management also empowers us to learn from the past. AI systems, much like humans, are not perfect—they make mistakes, sometimes with severe consequences. By documenting these mistakes, analyzing them, and making that knowledge available, we can create an environment of continuous improvement. We must also be willing to ask ourselves the hard questions: Are the decisions made by our AI systems transparent, and explainable? Are we accountable for their outcomes? Who benefits, and who bears the cost? Ethical AI development demands that we confront these questions openly, with a commitment to ensuring just, trustworthy, and responsible AI solutions.

As we move from data to justice, it is crucial that we uphold the principles of fairness, transparency, and accountability. Ethical AI is not just about mitigating harm—it is about proactively creating systems that uplift and empower all people. It is about leveraging AI to reduce inequalities, promote access to opportunities, and enable better decision-making across all organizations, companies, government, and society. In an era where artificial intelligence continues to reshape our world, it is our responsibility to ensure that its development serves all of humanity fairly, equitably, and justly. By incorporating robust knowledge management practices, we build a foundation that allows us to achieve these goals.

Summary and Points to Consider

• AI’s Economic Impact: An investment of $3 trillion in AI is expected to contribute $16 trillion to the global economy by 2030. Large language models (LLMs) have been identified as the fastest-adopted technology, influencing numerous sectors.
• Organizational Adoption of AI: 80% of Fortune 1000 companies will integrate AI into their products/services in the next 12 months.
• AI Delivery of Knowledge: AI is set to transform how knowledge is delivered and accessed. It’s expected to predict trending knowledge areas, personalize knowledge delivery based on user preferences, and improve content decisions using machine learning.
• Ethical Challenges of AI: It is important to use unbiased data and algorithms. AI applications need to be explainable, trustworthy, and ethically sound.
• People-Centered AI Approach: We need to advocate for AI applications that support inclusivity, well-being, and human-centered values. It emphasizes the importance of transparency, safety, and accountability in AI design and implementation.
• It is crucial that we uphold the principles of fairness, transparency, and accountability. Ethical AI is not just about mitigating harm—it is about proactively creating systems that uplift and empower.