DATA_FAIR 2025: Reflections on Inclusive AI, Bias, and Ethical Innovation

The DATA_FAIR Conference 2025 in Ljubljana, Slovenia, made history with an unprecedented lineup of women speakers in data engineering, data science, and machine learning—a first for Slovenia. While amplifying female voices was a central theme, the event fostered an inclusive space for participants of all genders, identities, and technical backgrounds from across the globe to engage in critical discussions about AI and data-driven technologies.

With a key focus on mitigating bias in AI training datasets, the conference explored some of the most pressing challenges and exciting possibilities in the field.

The broad themes covered in the DATA_FAIR 2025 include:

• Practical applications of data engineering technologies
• Current and emerging trends in machine learning (ML) and artificial intelligence (AI)
• Ethical considerations in data engineering

P.C: Gaja Dornik

Key Highlights from DATA_FAIR 2025

Discussions on Bias and AI Ethics

A major theme throughout the conference was the uncomfortable but necessary conversation around bias, ethics, and accountability in AI. The discussions weren’t just theoretical—they dug into the real challenges of building AI that is fair, transparent, and actually serves people rather than reinforcing existing inequalities.

One of the biggest issues? Bias in data. It’s everywhere, shaping the way AI models behave in ways that are often invisible until the damage is done. The focus wasn’t just on spotting these biases, but on practical ways to build more inclusive, representative datasets—because an AI system is only as fair as the data it’s trained on.

But better data alone isn’t enough. Engineering best practices matter just as much. The way data pipelines are designed determines whether biases get amplified or corrected. Conversations highlighted the need for rigorous data quality checks, diverse testing environments, and transparent decision-making processes—essential elements for AI that people can actually trust.

Then there’s the question of AI in high-stakes industries like healthcare and finance. When AI is flagging fraudulent transactions or detecting early signs of cancer, mistakes don’t just lead to inconvenience—they can have life-altering consequences. The ethical responsibility here is massive, and there’s still a long way to go in ensuring that AI-driven decisions are reliable, explainable, and free from hidden biases.

That’s where explainable AI (XAI) comes in. If an AI system rejects a loan application, diagnoses a disease, or denies insurance, people deserve to know why. Yet too often, AI remains a “black box,” offering no insight into how it reaches its conclusions. The push for greater transparency and accountability was clear—because trust in AI isn’t just about accuracy; it’s about whether people can understand and challenge its decisions.

Lastly, benchmarking AI against ethical standards is no longer optional. Industry-wide guidelines for fairness, transparency, and responsible deployment aren’t just a regulatory checkbox—they’re the foundation for an AI future that doesn’t repeat the mistakes of the past.

At the heart of all these conversations was a simple but urgent truth: AI is never neutral. The way we build and deploy it reflects our choices, values, and priorities. Getting it right isn’t just a technical challenge—it’s a societal one.

Conversational AI: Charting the Evolution

A standout session was delivered by Amisha Kothari and Ana Petrova,System Engineers at DevRev on Conversational AI: From Chatbots to AI Agents. The presentation covered:

• The Evolution of Conversational Technologies: Tracing the journey from basic chatbots to sophisticated AI agents, highlighting their increasing role in human-computer interactions.
• Challenges in Conversational Systems: Addressing common pitfalls such as security vulnerabilities and embedded biases in AI-powered communication.
• Future Applications of Conversational AI: Envisioning AI agents as collaborative tools that enhance productivity and decision-making, while maintaining ethical integrity.

This session sparked thought-provoking conversations on designing conversational AI systems that prioritize both utility and fairness.

P.C: Gaja Dornik

DATA_FAIR Café: Fostering Collaboration and Knowledge Sharing

A standout feature of DATA_FAIR 2025, the DATA_FAIR Café provided an interactive space for meaningful discussions at the intersection of AI, data science, and society. Moving beyond traditional presentations, it fostered small-group, expert-led conversations where attendees could engage directly with peers and specialists, exchanging ideas and exploring emerging challenges in an informal yet intellectually rich setting.

P.C: Gaja Dornik

The Café encouraged multidisciplinary discussions, bringing together diverse perspectives to examine both technical advancements and broader implications. Topics ranged from AI-driven automation in engineering to the ethical concerns of biased data and the environmental impact of large-scale AI models. Some of the key conversations included:

• AI Assistants as Data Scientists or Engineers: How tools like ChatGPT, Claude, and Copilot are transforming workflows.
• At the Bottom of the AI Rabbit Hole: Horror or Wonderland? – A gamified exploration of AI’s risks and rewards.
• Data-Facilitated Violence: Examining the impact of biased data and the ethical responsibilities of data professionals.
• Data-Driven Desire & AI in Relationships: Understanding the role of AI in shaping modern relationships.
• Gender and Color in Data Visualization: Analyzing how design choices influence perception and inclusivity.
• The Green Elephant in the Data Center: Addressing the environmental footprint of AI and advocating for sustainable data practices.

By facilitating open dialogue, the Café emphasized the need to approach AI and data science with both innovation and responsibility. The discussions underscored how technical decisions are deeply intertwined with ethical considerations, reinforcing the importance of critical thinking and cross-disciplinary collaboration in shaping the future of technology.

As AI continues to evolve, spaces like these remain essential—not just for staying informed, but for actively contributing to a more thoughtful, responsible, and impactful technological future.

P.C: Gaja Dornik