QCon London 2026: Ethical AI Is an Engineering Problem

2026-03-24 · Source: InfoQ · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Responsible AI Engineering · Depth: Intermediate, short

Summary

At QCon London 2026, Clara Higuera, Responsible AI Program Lead at BBVA, presented on treating ethical AI as an engineering problem rather than solely a governance issue. The session highlighted how AI system failures, such as the wrongful arrest caused by facial recognition misidentification, stem from technical choices like unrepresentative training data or poor evaluation pipelines. Higuera emphasized that ethical properties like fairness, transparency, security, sustainability, and accountability must be integrated throughout the AI lifecycle, from data collection and model architecture to deployment. This involves evaluating datasets for representativeness, measuring model behavior across demographic groups, and ensuring system observability, akin to how other critical industries developed safety standards.

Key takeaway

For AI Architects and Machine Learning Engineers building critical AI systems, you should proactively integrate ethical considerations as core engineering requirements. Incorporate fairness evaluations, explainability analyses, and security testing throughout your development lifecycle to mitigate risks like bias and security vulnerabilities. This approach helps ensure systems are robust and socially responsible, especially as formal standards are still evolving.

Key insights

Ethical AI is an engineering challenge requiring integration of ethical checks throughout the development lifecycle.

Principles

• Ethical properties are measurable engineering requirements.
• AI systems encode values embedded in their design.
• Technology often evolves faster than regulation.

Method

Embed ethical checks throughout the AI development lifecycle, including fairness evaluation during training, explainability analysis pre-deployment, security testing, and production monitoring for unexpected behavior.

In practice

• Evaluate datasets for representativeness.
• Measure model behavior across demographic groups.
• Test against adversarial attacks like prompt injection.

Topics

• Ethical AI
• Responsible AI
• AI System Design
• Algorithmic Bias
• AI Lifecycle Management

Best for: AI Architect, Machine Learning Engineer, Computer Vision Engineer, AI Engineer, MLOps Engineer, AI Ethicist

Related on AIssential

• AI as a Double-Edged Sword: Opportunities and Ethical Challenges — AI's dual nature demands responsible development, balancing innovation with ethical safeguards like transparency and human oversight.
• Towards Responsibly Non-Compliant Machines — Engineering AI agents to responsibly refuse user requests requires defining justifications, override mechanisms, and managing risks.
• Top AI ethics and policy issues of 2025 and what to expect in 2026 — 2025 marked AI's shift from testing to deployment, bringing both regulatory enforcement and new ethical challenges.
• The Good Robot podcast: the role of designers in AI ethics with Tomasz Hollanek — Design is central to AI ethics, embedding values and requiring critical approaches for just human-AI interaction.
• Trustworthy AI Software Engineers — Trustworthy AI software engineers must integrate into human-AI teams, extending beyond coding to socio-technical tasks and ethical alignment.
• Is Your AI Unfair? Why Responsible AI is the New Non-Negotiable in Customer Experience — Biased AI, reflecting its training data, can lead to discriminatory outcomes and significant business liabilities in customer experience.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by InfoQ.