Can We Trust AI? Taja Hillier on Data, Oversight, and Aviation’s AI Future

In this Inside Aviation interview filmed at EATS 2025, Halldale Correspondent Atul Chandra speaks with Taja Hillier, Chief Data & AI Officer and co-founder of Mission Decisions, a UK-based data and AI services company with roots in defence.

Hillier brings deep expertise in IT, statistical methods, and economics, combined with experience in safety-critical environments. 

Her message is clear: AI trust starts with data and transparency: 

“You can’t have AI strategy without good data. AI is still a model - without proper preparation, it fails.”

The discussion explores why aviation companies often struggle with AI adoption. Hillier warns against “plug-and-play” approaches and stresses the need for data scientists and engineers to prepare structured and unstructured data correctly. 

She highlights common pitfalls, such as misunderstanding retrieval-augmented generation (RAG) and neglecting transparency: “If a vendor can’t explain how their AI works, don’t trust them.”

Hillier also addresses misconceptions about AI as a “black box,” emphasising that companies can and should understand why models produce certain outputs. Even when AI is implemented correctly, human oversight remains essential: “First runs may look plausible, but responsibility always lies with people.”

Looking ahead, Hillier predicts AI will become enterprise-ready within two years, provided companies prioritise accuracy, scalability, and regulatory compliance. 

Her advice for organisations:

• Employ skilled data scientists and engineers.
• Ensure transparency and traceability.
• Maintain human-in-the-loop accountability.

Watch the full interview to learn why data-driven discipline and oversight are key to making AI work for aviation.