How AI can transform insolvency processes for firms in the digital age

Sometime last year, we walked into a small office in Nairobi, facing an all too familiar situation, another business in distress.

A once-thriving fintech company had officially gone into administration. Despite the best efforts of its directors, a tough economy, regulatory challenges, and dwindling investor confidence left no option but to bring in insolvency practitioners to take over.

For those unfamiliar with the administration process, it is a structured approach aimed at stabilising a financially distressed company by either turning it around or ensuring the best possible return for creditors.

It is a high-pressure, high-stakes process, and like many before us, we knew the road ahead would be tough. What we didn’t anticipate was just how much manual, repetitive work we would have to do over the next few months. It made us wonder: Could Artificial Intelligence (AI) have made the difference?

One of the first statutory obligations we had to fulfil was to publish notices in newspapers and the Kenya Gazette, formally announcing the administration.

This required drafting the notices, getting them reviewed and approved, sending them to the publishers, and then waiting for confirmation that they had been printed.

Missing a deadline could mean legal non-compliance, so we found ourselves caught in a cycle of emails, phone calls, and constant checks to verify publication.

AI could have streamlined this process through workflow automation, ensuring that notices were automatically generated, reviewed, scheduled for publication, and tracked in real time, eliminating the need for manual follow-ups.

A critical step in the administration process involves identifying creditors and managing their claims. This means combing through vast amounts of transaction records to compile an accurate list of those to whom the company owed money.

Notifying them in the current manual setup requires drafting and sending numerous emails and letters and waiting for responses, following up with those that are unresponsive, and manually logging claims as they come in.

Some creditors may respond immediately, while others require multiple reminders, and still others may be untraceable.

AI can transform this process by scanning financial records to generate a real-time creditor list, automatically dispatch notifications, track responses, and flag duplicate or inconsistent claims, saving the administration team valuable time and resources.

Whilst in this case, the directors submitted a statement of affairs detailing the company’s assets, liabilities, and financial position, in an insolvency process, trust is never enough. Every figure must be verified. This meant analysing thousands of transactions, reviewing financial statements, and manually cross-referencing records to ensure accuracy.

Given that the company handled large volumes of financial data, this task was particularly time-consuming.

AI-powered auditing tools could have instantly analysed transactions, identified discrepancies, and flagged irregularities, allowing the team to focus on critical aspects of restructuring rather than spending long hours combing over spreadsheets.

With operations winding down, another challenge emerged: monetising the company’s assets. As a fintech company, the company’s most valuable resources weren’t physical inventory but rather its data, digital platforms, and customer relationships.

Determining how to extract value from these intangible assets was complex, requiring careful consideration of regulatory constraints, cybersecurity concerns, and potential buyers.

Instead of manually evaluating offers and negotiating each deal individually, AI-driven analytics could have assessed market trends, recommended optimal monetisation strategies, and even predicted potential buyers’ interest based on past transactions.

Without such tools, we had to rely on traditional valuation models and lengthy discussions to determine the best course of action.

As we navigated these challenges, it became clear that the company’s existing infrastructure was no longer sustainable. Maintaining its data centres and operational hubs was too costly, and a transition to a more streamlined setup was necessary.

Identifying alternative locations, negotiating lease terms, and managing the logistics of the move required intensive effort. AI could have optimised this process by forecasting cost savings, analysing staffing needs, and even automating asset relocation planning. Instead, we relied on time-consuming assessments, financial projections, and extensive negotiations to reach decisions.

This experience demonstrates that insolvency processes remain highly manual, repetitive, and inefficient. AI is not about replacing insolvency professionals with modern technology—it is about enhancing accuracy, efficiency, and speed.

From automating creditor notifications and claims validation to AI-driven audits that detect inconsistencies and workflow automation for legal compliance, the potential benefits are clear. With AI, real-time financial modelling could improve restructuring decisions, while smart asset monetisation strategies could maximise recoveries for creditors.

This case highlights the urgent need for the insolvency industry to embrace digital transformation. While the world moves towards automation and data-driven decision-making, many insolvency procedures remain entrenched in outdated, paper-based methods. Professionals in the restructuring, turnaround, and insolvency industry must embrace AI as an invaluable tool that can drive faster resolutions, optimised recoveries and smarter processes, which will drive better outcomes for distressed companies and their stakeholders.

- The writer is a Deal Advisory and Strategy Associate with KPMG Advisory Services Ltd; ashleyodhiambo@kpmg.co.ke. The views and opinions are those of the author and do not necessarily represent those of KPMG