Australia is in the middle of a business failure crisis by any historical measure. ASIC reported 14,722 companies entered external administration in 2024–25, a 33.2% jump on the prior year and the highest figure since 1999–2000. Personal insolvencies tied to business failure added another 3,536 cases, representing 78.8% of new liabilities despite being just 28.8% of new filings, according to AFSA.

That’s the field AI is entering, and it’s entering from several directions at once.

Catching trouble earlier

The most important thing AI could change about insolvency isn’t the wind-down. It’s the bit that comes before the wind-down, when a company is sick but not yet dead and there are still options on the table.

Directors and their advisors have always struggled with this. Cash flow feels fine until it doesn’t. A profitable business on paper can be technically insolvent. And the Corporations Act 2001 already imposes duties on directors to avoid trading while insolvent, but recognising that threshold in real time, without a forensic accountant on retainer, is genuinely hard.

Machine learning tools designed for financial analysis can now scan a business’s accounts, payment patterns, and creditor behaviour and flag deterioration early, before it becomes a statutory problem. A recent review in the International Insolvency Review noted that AI algorithms can process large datasets of financial information to identify companies at risk of insolvency, providing early warning to both directors and creditors.

That’s significant. The advice you’ll hear from any insolvency practitioner is that the earlier you seek help, the more options you have. Voluntary administration, restructuring, a deed of company arrangement, these paths stay open when there’s still cash and time. By the point a creditor petitions for winding up, most of those doors are already closed.

For small business owners specifically, the gap between “this feels hard” and “I need a liquidator” often goes unexamined for far too long. An AI-assisted accounting or banking platform that flags warning signs, sustained late payments, creditor pressure, deteriorating working capital, could be the equivalent of a check-engine light that most small businesses currently don’t have.

Inside the wind-down itself

When a company does enter external administration in Australia, the registered liquidator or voluntary administrator takes control and faces an immediate mountain of work: assembling creditor lists, reviewing transactions, preparing statutory reports for ASIC, tracking dozens of deadlines, and corresponding with hundreds of parties.

This is exactly the kind of work AI is good at.

Australian insolvency software company Ender has built an AI platform purpose-designed for Australian practitioners operating under the Corporations Act 2001 and the Bankruptcy Act 1966. Their system reads bank statements automatically, surfaces potential voidable transaction candidates, payments that might fall under the unfair preference provisions of Part 5.7B, and drafts statutory reports from documents already on file. Statutory deadlines are tracked automatically across the administration.

The legal point matters: AI can flag a transaction as a potential voidable dealing. It cannot void it. Only a court does that. The registered liquidator reviews every candidate and makes the call. But widening the coverage, scanning all transactions rather than sampling a proportion due to time constraints, means fewer things get missed.

Asset recovery is another area. A liquidator needs to identify and recover assets, including assets that may have been moved in the lead-up to failure. AI that can detect patterns across financial records, related-party transfers, unusual payments, timing relative to when debts arose, can do in hours what previously took weeks of manual review.

What creditors are getting

Creditors in Australian insolvency proceedings have historically been poorly served by information. They get a creditor report, an invitation to a meeting, and periodic updates, most of which arrive well after the fact.

AI is beginning to change what’s possible here too. Creditors, particularly institutional creditors like banks, the ATO, or trade creditors with large exposures, can now run their own analysis on a debtor’s financial position using AI tools, rather than relying entirely on what the administrator chooses to disclose. The ATO is already the largest single creditor in many insolvency matters, and tools that give it faster, cleaner insight into company finances could change how aggressively it moves and when.

For trade creditors and smaller businesses owed money, the shift is slower. But platforms that monitor the ASIC register, flag when a customer or debtor enters administration, and automatically pull the relevant practitioner contact details are already available. The question is whether enough SMEs are using them.

The disruption side of the equation

The Mondaq analysis from Melbourne-based insolvency practitioners Clifton Hall raises a less comfortable point: AI isn’t just changing how insolvency is managed. It may be driving more of it.

The argument is straightforward. Businesses that rely on repetitive, process-driven work, administrative services, document preparation, routine professional tasks, are under growing margin pressure as AI tools reduce the cost of producing the same output. When everyone in an industry can access the same technology, the competitive advantage disappears and all that’s left is price pressure. Businesses running on thin margins before AI hit that wall faster.

Professional services firms are specifically flagged: accounting, legal, consulting. Tasks that previously took a junior professional hours, drafting correspondence, reviewing contracts, preparing reports, can now be completed in minutes. Clients notice. Billing models built on hourly rates for knowledge work are already being questioned. Smaller practices, without the financial runway to absorb a pricing transition, face the hardest adjustment.

This isn’t hypothetical. The same disruption pattern played out with digital retailing and media. Businesses that didn’t adapt quickly enough failed. Insolvency practitioners saw it directly. The question with AI is whether the disruption is faster and broader, and the current data suggests it probably is.

What this means if you run a small business

There are two things worth taking from this.

First, watch your margins carefully, not just your revenue. A business can look healthy on top-line numbers while its cost structure is being quietly hollowed out by competitors who’ve adopted AI more aggressively. If you’re in a service business where AI tools are becoming standard, and you haven’t looked at your pricing model recently, that’s worth doing now rather than in six months when the pressure is harder to ignore.

Second, if you’re feeling financial stress, get advice early. Under the Corporations Act, directors have personal liability exposure for trading while insolvent. The voluntary administration regime under Part 5.3A exists precisely to give companies breathing room before full liquidation becomes the only option, but it only works if there’s still something to work with. The AFSA data is clear that business-related insolvencies represented 78.8% of new liabilities in 2024–25 despite being under a third of filings. The size of losses when businesses fail suggests many owners are waiting too long.

AI may eventually make early warning more automatic, a platform that flags when your business is heading toward trouble, the way your bank already flags unusual card transactions. That’s not widely available to SMEs yet. Until it is, the responsibility for watching those numbers stays with the director.

📬 The SmallBizAI Brief

One practical AI tip for Australian small business. Every Tuesday. Free.

Join business owners getting smarter about AI every week.

Subscribe free →

Sources

📬 Subscribe to our weekly newsletter | 💼 Free resources & AI prompt packs

📖 More deep dives: News Deep Dives

📊 Compare AI tools side by side | 💼 Free resources & AI prompt packs

: News Deep Dives →