AI Promised Efficiency; Now It Is Creating Commercial Disputes

However, as adoption increases, so too does the potential for commercial disputes.

There is a risk to businesses who have embraced AI without giving full consideration to who bears responsibility when the technology produces inaccurate results, uses incorrect data, fails to perform as expected, creates regulatory compliance issues or who is responsible for human oversight. 

As a result, disputes are emerging that are not necessarily about the technology itself, but about the contracts, expectations and risks surrounding its use.

For business owners, directors and senior managers, understanding these risks now could prevent expensive litigation later.

Where Are AI-Related Disputes Arising?

Although AI is often presented as a technological issue, many disputes ultimately come down to familiar legal principles.

Businesses are finding themselves involved in disagreements concerning software implementation projects that fail to deliver promised results, inaccurate outputs relied upon for business decisions, ownership of AI-generated content and breaches of confidentiality arising from information entered into AI platforms.

The common theme is expectation and contractual terms, policies and procedures not matching those expectations.

Many organisations have purchased AI-based products or services based on ambitious claims regarding efficiency, cost savings or performance improvements. When those expectations are not met, disputes can quickly arise.

The Contract Matters More Than Ever

One of the most important lessons from technology disputes is that the contract remains the primary source of protection.

Many AI products are being adopted rapidly through standard terms and conditions, software licences or online subscription agreements that receive little scrutiny during procurement.

Unfortunately, those documents often contain extensive limitations of liability, broad exclusions of responsibility and carefully drafted disclaimers regarding the accuracy of outputs.

Businesses frequently assume that a supplier will be accountable if an AI system gets something wrong. The contractual position may be, and usually is, very different.

Before implementing AI solutions, organisations should consider:

• What exactly is the supplier promising?
• Are specific outcomes being guaranteed?
• Who owns any content generated?
• What happens if the system makes an error?
• Are liability caps appropriate?
• Does the contract adequately address confidentiality and data protection concerns?

These are increasingly becoming dispute resolution questions rather than procurement questions.

Human Oversight Remains Critical

Another emerging source of disputes involves overreliance on AI-generated information.

Businesses understandably seek efficiencies, but AI systems can produce inaccurate or misleading outputs. This risk is commonly referred to as "hallucination", where an AI system presents incorrect information with apparent confidence.  Also, AI systems may present aspects of output which are correct, but the AI system has made assumptions or not taken into account key factors, meaning the output is wrong for the particular circumstances.

Where commercial decisions are made solely on the basis of AI-generated content, without appropriate checks and human oversight, difficult questions can arise regarding responsibility for resulting losses.

From a dispute resolution perspective, organisations should ensure that appropriate governance structures remain in place. Human review processes, clear approval procedures, clear AI policies and documented decision-making can all play an important role in reducing both risk and potential liability.

Importantly, these safeguards may also provide valuable evidence if a dispute subsequently arises.

Record Keeping Could Become Your Best Defence

Evidence remains at the heart of any commercial dispute.

Historically, disputes often revolved around emails, meeting notes and contractual documents. Increasingly, organisations may also need to consider records relating to AI-generated outputs and the prompts used to create them.

Businesses should think about what information was provided to the AI system, how outputs were reviewed, who authorised decisions based upon those outputs and what contractual protections were in place at the relevant time.

Clear records can help establish what happened and why.

Conversely, poor governance may make it significantly more difficult to defend claims or pursue recovery against suppliers.

The Regulatory Position Is Still Evolving

Adding another layer of complexity is the developing regulatory landscape.

The legal framework governing AI continues to evolve both domestically and internationally. While legislation and guidance are developing, businesses cannot afford to wait for complete regulatory certainty before addressing risk.

Commercial disputes rarely arise because organisations intentionally ignore legal obligations. More often, they arise because risks were not identified early enough.

Companies that proactively review contracts, data governance arrangements and internal policies are likely to be in a stronger position than those that adopt AI first and consider legal consequences later.

Also, companies that do not fully consider the position around protecting legal privilege for any legal advice received by checking the advice via an AI system or who turn to AI (especially public AI systems) for initial advice on the legal position can compound risks.  In a worst-case scenario, all those inputs and outputs of the AI system may be evidence which is disclosable to the other party to the dispute.  Extra emphasis on those aspects in AI policies and procedures can at least help prevent a dispute becoming even worse.

What This Means for Businesses

Artificial intelligence undoubtedly offers significant opportunities. For many organisations, it is already delivering efficiencies and competitive advantages that would have been unimaginable only a few years ago.

However, from a dispute resolution perspective, AI is not creating entirely new legal problems. Instead, it is creating new scenarios in which familiar legal issues emerge.

Contracts, governance, accountability and evidence remain fundamental.

Businesses considering the use of AI should treat legal risk assessment as part of implementation rather than an afterthought. Reviewing contractual arrangements, maintaining appropriate AI policies, human oversight and ensuring effective record keeping can significantly reduce the likelihood of disputes arising.

Where disagreements do occur, organisations that have taken these steps will generally find themselves in a much stronger position.

The growth in AI adoption, coupled with increasing concern from regulators, insurers and businesses, is creating new opportunities for commercial disputes. In practice, many technology disputes now involve AI-enabled products or services, raising familiar legal issues concerning contractual performance, liability, responsibility, intellectual property and data governance.

As AI continues its rapid integration into everyday business operations, the most successful organisations are likely to be those that strike the right balance between innovation and effective risk management.