New AI Technology Tackles Hallucinations with Uncertainty Awareness

BREAKING: A groundbreaking artificial intelligence technology aimed at drastically reducing hallucinations in AI-generated responses has been unveiled, according to a report from Sky News on October 14, 2025. Dubbed the Uncertainty Engine, this innovative AI utilizes ‘trust cues’ to enhance user confidence, paving the way for its adoption in critical sectors such as energy and healthcare.

This new breed of AI, identified as uncertainty-aware AI, addresses the persistent issue of hallucinations—misleading outputs produced by algorithms when they encounter incomplete or inconsistent data. Rather than delivering absolute answers, this technology provides informed assessments based on existing evidence while clearly acknowledging uncertainties. This dual approach allows human operators to evaluate the reliability of the information and seek further data when necessary.

The Uncertainty Engine is being developed by DigiLab, a start-up focused on designing next-generation compact nuclear reactors intended for use in proximity to residential areas. Company founder Tim Dodwell emphasized the critical nature of this technology in transformative applications, stating,

“I see no real future for AI unless we answer another problem: can we trust it? Human lives are at stake and people have to take responsibility for what the algorithms will do.”

Currently, the Uncertainty Engine is undergoing testing in practical scenarios, including open-sea navigation. The AI assistant is engineered to assist sailors in charting the fastest and safest routes by analyzing vast datasets on wind conditions, sea states, boat performance, and competitor positions. By leveraging this information, sailors can make data-driven decisions, enhancing overall operational safety.

Beyond its immediate applications in maritime navigation, uncertainty-aware AI holds promise for significant advancements in the shipping industry. The technology is expected to optimize routes, taking into account variable weather conditions and potential disruptions, such as canal closures. For instance, even a modest 5% reduction in fuel consumption in the UK shipping sector could equate to removing hundreds of thousands of cars from the roads, showcasing the potential environmental benefits of this innovative technology.

As the world increasingly adopts AI in high-stakes environments, the implications of managing uncertainty become ever more critical. The Uncertainty Engine represents a pivotal step forward in building trust and reliability into AI systems, ensuring they can be safely integrated into sectors where human lives and the environment are at stake.

Stay tuned for further updates on this developing story as the Uncertainty Engine prepares for wider implementation across various industries.