Although AI adoption appears ubiquitous, returns remain elusive

The reality is that too many AI systems never cross from pilot into production. What holds them there is seldom capability – it is trust – which if designed in with intent, allows the value to convert into commercial return.

In McKinsey’s 2026 AI Trust Maturity Survey, close to two-thirds of organisations named security and risk concerns as their main barrier to scaling agentic AI, ahead of both regulatory uncertainty and technical limitations. The binding constraint is not the capability of the technology but the confidence to deploy it.

The same research finds trust increasingly regarded as a business enabler rather than a compliance exercise, with adoption now driven by value and performance rather than by regulation alone. Trust does two things at once: it allows an organisation to realise value from its AI by supporting sustained adoption, and it allows it to manage a widening landscape of risk as systems take on greater autonomy.

Trust is the precondition for return, not a reward for it

Trust is not a by-product of commercial success, acquired once a product is already selling. It is the condition that must be met before the value in an AI system can be released at all. The capability is usually present in the pilot, what is missing is the assurance that lets a buyer purchase it, a board approve it into production, a regulator permit it and an insurer underwrite it. Each of those is a commercial gate, and trust is what opens it. Withhold it, and the value stays stranded, however capable the model.

The relationship is measurable. McKinsey reports that organisations making substantial investment in responsible-AI practices achieve both higher trust maturity and a materially greater likelihood of realising financial benefit. This includes earnings impact above five percent, and frames that investment as a driver of value rather than a constraint on innovation. Trust is therefore necessary, though not on its own sufficient: a well-governed AI product that no one wants will still fail.

For the large majority of organisations whose AI works and yet earns little, the missing factor is not a better model. It is the demonstrable trustworthiness that turns a capable system into an adopted one.

The dimensions of trust, and the instruments that build it

Trust of this kind is not achieved by a single act of compliance. It has several dimensions, whether an AI system is safe and reliable, whether it is fair and free from unlawful bias, whether its outputs can be explained, whether it is accountable to identifiable owners, and whether all of this can be evidenced to someone outside the organisation.

No single instrument addresses them all, and regulation is only one of the available routes. The EU AI Act makes certain of these obligations mandatory for the European market. Beyond it sits a range of instruments an organisation can adopt voluntarily, and for commercial reasons: the ISO/IEC 42001 standard, which provides a certifiable AI management system; the NIST AI Risk Management Framework, an internationally recognised structure for identifying and managing AI risk, and principle-based frameworks, notably the OECD AI Principles and the 12 Principles of AI governance that Anekanta® developed into a commercial context. These translate high-level values into decisions a board can actually take. Used together, these are how an organisation evidences, across every dimension a customer, regulator, insurer or investor cares about, that its AI can be trusted. This article considers the regulatory position first, because its timetable has just changed, before looking at how the wider set of instruments can be put to work.

The regulatory clock has moved – the commercial one has not

Of these instruments, the EU AI Act is the one that carries the force of law, and it is the one whose timetable has just shifted. On 29 June 2026, the Council gave its final green light to the Digital Omnibus package which defers the Act’s high-risk obligations for standalone systems to 2 December 2027. Although that adjusts the regulatory schedule it does not adjust the commercial one. Customers, procurement teams, insurers and investors are scrutinising AI more closely now than a year ago, and their expectations are not set in Brussels.

The organisations that read the deferral as a reprieve will reach the new deadline no more trusted than they are today. Those that treat the interval as what it is, the time it takes to build and evidence a governed AI capability, will reach it demonstrably trustworthy, having used the period to convert that trust into contracts, cover and capital.

An AI management system takes many months to establish and certify. The period to December 2027 is, in practical terms, the window in which it can be built without disruption to the business.

How trust is built in practice

Building trust across those dimensions is a disciplined exercise, and it is the work Anekanta® undertakes with boards and senior leaders. It begins with knowing what the organisation is actually running, identifying AI use cases, classifying them against the EU AI Act, and assessing the risks and inferences they create, including the indirect inferences that AI systems draw from proxy data. This is the foundation of any credible trust claim, and it also directs investment toward the use cases most likely to produce value and away from those that carry disproportionate exposure.

It then requires a means of proving trustworthiness to others. An ISO/IEC 42001 management system, informed where useful by the NIST framework and anchored in clear governing principles, converts internal practice into external, portable evidence, of the kind a customer’s procurement function, an insurer or a regulator will accept in place of assurances. Established once, it is reusable across every market and transaction, and it embeds governance across the development and deployment lifecycle rather than adding it after release.

Finally, it depends on competent accountability at board level. Governance ultimately rests with a board that understands what the organisation is standing behind and is equipped to decide on it, which is why AI literacy among senior leaders, also a legal requirement under Article 4 of the EU AI Act, is what makes that accountability real. A board able to weigh AI opportunity against AI risk with confidence is itself a signal of trustworthiness to customers, investors and regulators.

The commercial reward

On the evidence, the organisations that will realise the commercial reward from AI are not those with the most advanced models. They are those that have made their AI trustworthy enough to be adopted, deployed and relied upon, and can prove it. Trust is what moves an AI system out of the pilot and into production, clears it through a customer’s procurement process, and allows a business to scale it across regulated markets with confidence. Governance is how that trust is built, and, built well, it is not a cost of doing business but the precondition of the commercial case for AI.

How Anekanta® can help

Anekanta® helps boards, CEOs and senior leaders turn AI ambition into trustworthy, commercially viable adoption, working across AI strategy and use-case evaluation, EU AI Act risk assessment and classification, AI literacy for senior leaders and across enterprises, and preparation for ISO/IEC 42001. Explore our services further.

To build board-level competence and confidence, our AI Literacy and Governance Workshops move from principles to practice. The next open session is hosted by the ISACA London Chapter on 23 September 2026.

To assess where your organisation’s AI portfolio stands, or to prepare for ISO/IEC 42001, begin a specialist advisory enquiry with our team.

Find Out How to Build Trust in AI and Accelerate Your Organisation’s Success

Intellectual Property: © 2016–2026 Anekanta®. All rights reserved. Unless otherwise expressly stated, all materials published on this website, including the Anekanta® AI Governance Framework for Boards, the 12 Principles, and all AI risk and impact evaluation methodologies, software, models, diagrams, text and materials, are proprietary intellectual property of Anekanta®. No reproduction, adaptation, distribution, or commercial exploitation is permitted without prior written authorisation. No rights are granted other than those expressly stated. The Anekanta® AI Governance Framework and 12 Principles are developed, maintained and continuously enhanced as part of Anekanta®’s proprietary governance architecture.
Professional Disclaimer: The information provided on this website is for general informational purposes only and does not constitute legal, regulatory, financial or professional advice. Any reliance placed on the information is strictly at the user’s own risk. Professional advice should be sought in relation to specific circumstances through a formal engagement with Anekanta®.
Use of Generative AI: Generative AI tools may be utilised in research and drafting processes. All published materials are subject to substantive human review, professional judgment and oversight prior to release.