AI Doesn't Have an Intelligence Problem. It Has a Trust Problem

Artificial intelligence has rapidly become one of the defining business conversations of the decade. Much of that discussion has centred on increasingly capable foundation models, from generative AI and copilots to ever-more sophisticated reasoning engines. The underlying assumption has been that as the technology becomes smarter, enterprise adoption will naturally accelerate. 

That may prove true for low-risk applications, but industrial environments present a fundamentally different challenge. The question facing asset-intensive organisations has never been whether AI can generate an answer. It is whether that answer can be trusted when it influences decisions involving operational continuity, capital allocation or critical infrastructure. 

There is an enormous difference between asking AI to summarise a report and asking it to recommend an inventory purchase worth millions of dollars, or to determine an appropriate maintenance strategy or identify the component required to keep a mine, manufacturing facility or utility operating safely and efficiently. 

As AI begins influencing decisions with significant financial and operational consequences, intelligence alone is no longer enough. Trust, governance and accountability become just as important as the technology itself. 

This distinction is becoming increasingly relevant as industrial organisations accelerate investment in digital transformation. Poor data quality costs organisations an average of US$12.9 million every year, according to Gartner research, while organisations continue investing heavily in enterprise software and AI to improve productivity, optimise inventory and reduce operational risk. Yet many businesses are attempting to apply increasingly sophisticated AI models to operational data that has accumulated over decades with little consistency, governance or standardisation. 

Across mining, manufacturing, energy and utilities, it is common to find duplicate material records, inconsistent naming conventions, obsolete supplier information, incomplete specifications and fragmented maintenance histories spread across multiple enterprise systems. These challenges existed long before artificial intelligence entered the conversation, but AI has the potential to amplify them rather than solve them. 

Technology cannot create certainty from unreliable information. It simply processes poor-quality data faster and produces recommendations that appear increasingly confident, regardless of whether the underlying information is accurate. 

This is why the enterprise AI conversation needs to move beyond model capability and towards decision architecture. Organisations should be asking where operational data originates, how it is governed, what controls exist before recommendations influence procurement or maintenance decisions, and where accountability ultimately sits when AI becomes part of the decision-making process. 

The most successful industrial AI implementations will not be those that remove people from decision-making. They will be those that combine trusted operational data with clearly defined governance frameworks and experienced professionals who remain accountable for outcomes. Human expertise has been built over decades through operational experience, engineering judgement and practical understanding of complex industrial environments. AI should strengthen that expertise by making it more accessible, more consistent and more scalable across an organisation, rather than attempting to replace it altogether. 

This human-in-the-loop approach represents a more sustainable model for enterprise AI because it recognises that technology and experience perform different but equally important roles. AI excels at analysing vast quantities of operational data, identifying patterns and surfacing recommendations at a speed no individual could achieve. Experienced engineers, planners and procurement specialists provide the context, judgement and commercial understanding needed to validate those recommendations before critical decisions are made. Together, they create far better outcomes than either could achieve independently. 

This shift is also changing the way enterprise software is evaluated. Organisations are placing increasing emphasis on governance, explainability and transparency alongside technical capability. They want to understand how recommendations are generated, what business rules have been applied, how operational data is validated and where human oversight remains embedded within the workflow. These capabilities are rapidly becoming prerequisites for enterprise AI adoption because buyers recognise that confidence in a recommendation is just as important as the recommendation itself. 

There is little doubt artificial intelligence will reshape industrial operations over the coming decade. The organisations that realise the greatest commercial value, however, are unlikely to be those deploying the largest models or automating the greatest number of decisions. They will be those that build systems capable of combining the analytical power of artificial intelligence with trusted data, disciplined governance and the operational judgement of experienced people, ensuring technology enhances human expertise rather than attempting to replace it. 

Ultimately, the future of enterprise AI will not be determined solely by advances in intelligence. It will be defined by the organisations that earn trust by combining technology with human judgement, creating decision-making environments where AI supports better commercial outcomes because experienced people remain at the centre of the process.


About Lexin Solutions 

Lexin Solutions is an Australian-founded, U.S headquartered enterprise software company helping asset-intensive industries optimize inventory, improve asset performance and unlock working capital through AI-powered supply chain intelligence. Working with organisations across mining, oil & gas, utilities and heavy industry, Lexin transforms complex operational data into actionable insights that reduce risk, improve decision making and deliver measurable commercial outcomes. 

About Derek List 

Derek List is the Head of Product and Co-Founder of Lexin Solutions, where he leads the strategy and development of MiLi, the company's AI-powered SaaS platform helping industrial organisations optimise indirect material supply chains. 

With a career spanning analytics, product strategy, enterprise software and supply chain transformation, Derek specialises in turning complex operational challenges into practical, scalable technology solutions. Working closely with customers across mining, energy, utilities and manufacturing, he is helping organisations improve inventory performance, strengthen governance and unlock measurable commercial outcomes through enterprise AI.