Air Safety at Risk: No AI Takeover in Aviation?

A new survey by the European Union Aviation Safety Agency (EASA) shows that aviation professionals remain cautious about artificial intelligence. While they see potential, many still worry about risks, accountability, and safety.

Survey Findings

The survey, conducted in January 2024 with 231 participants, explored comfort, trust, and acceptance of AI in aviation. The average acceptance score was 4.4 out of 7, showing moderate support but with reservations.

Nearly two thirds of respondents rejected at least one AI scenario. This reflects doubts about how far AI should go in critical aviation operations.

Key Concerns

The main concerns were limits of AI performance, data protection, privacy risks, and accountability for errors. Safety implications were seen as critical.

Respondents also feared over reliance on AI could cause de skilling. Human pilots and engineers may lose essential knowledge if AI dominates decision making.

Regulatory Oversight

Many stressed the need for strict oversight by EASA and national aviation authorities. They want regulations that ensure safety before AI gains wider trust.

EASA’s Guillaume Soudain said AI can improve efficiency and safety, but trust is key. He called for balanced rules that allow innovation without lowering standards.

European Commission’s View

Christine Berg from the European Commission underlined AI’s role in traffic flow management, predictive maintenance, and autonomous systems. She stressed that aviation must remain safety critical. AI systems, she said, must be explainable, reliable, and certifiable.

Industry Developments

At the AI Days conference, EASA showcased research on AI assurance, human factors, and ethics based assessment. Agencies such as the FAA and EUROCAE also discussed collaboration.

Industry projects included Boeing’s auto taxiing system, Lufthansa’s troubleshooting assistant, and SESAR programmes like JARVIS, DARWIN, and SynthAIR. These highlight the pace of aviation innovation.