76% of Australian businesses hit by AI failures, BSI finds

By Rebecca Stone

A BSI study published on Friday, November 6, 2026, found that 76% of Australian businesses had experienced at least one AI-related failure in the previous 12 months.

The result matters because BSI found recurring failures across live business systems, while only 37% of Australian organizations reported a formal AI governance program.

BSI produced the study with Alma Economics. It combined economic modelling with a survey of 1,287 business leaders across nine countries.

The study estimated that unmanaged AI risks could place between 0.1% and 3.6% of annual market revenue at risk across exposed sectors, depending on the type of failure.

Why are AI failures becoming routine?

BSI's Australian results describe repeated breakdowns rather than isolated incidents. Among businesses that reported a failure, 88% said they had suffered multiple incidents in the previous 12 months, while 9% recorded between 51 and 100 separate failures.

The study found the failures were not confined to trial projects. It recorded three main areas of operational impact:

• Internal workflow tools accounted for 34% of reported failures.
• Customer-facing systems accounted for 21%.
• Failures disrupted internal and external applications at the same time in 28% of cases.

Among Australian organizations that had encountered failures, BSI recorded a broad range of operational and financial effects:

• 50% reported downtime or efficiency losses, above the study's 44% global average.
• 35% faced regulatory or compliance consequences, compared with 29% globally.
• 31% reported security breaches.
• 25% recorded safety failures.
• 24% reported direct financial losses.
• 22% reported reputational damage.

BSI's categories extend the risk beyond technical reliability. The same failures produced downtime, compliance exposure, security incidents and direct losses across a majority of affected organizations.

How are businesses spending despite the failures?

Deployment continued despite the incident rates. The BSI survey found that companies were increasing their AI budgets and expected further increases.

• 72% of Australian businesses had seen year-on-year AI costs increase.
• 77% expected AI budgets to rise further.
• 50% said AI spending had diverted funding from other strategic initiatives.
• 81% had seen tangible operational benefits from AI.
• 72% reported a positive return on investment.
• 77% planned to expand their AI investment.

These results place AI governance inside the budget decision. The same companies report positive returns and planned expansion alongside downtime, compliance exposure and funding diverted from other initiatives.

What is the governance gap?

BSI highlighted a gap between adoption and oversight. Formal controls lagged behind the deployment of specific policies and risk-assessment systems.

• 53% of Australian organizations had policies covering the safe and ethical use of AI.
• 59% had standardized risk-assessment frameworks.
• Only 37% had established a formal AI governance program.
• Just 29% were taking steps to align with ISO/IEC 42001, the international standard for AI management systems. That share trailed China at 37% and the global average of 31%.
• Only 38% knew which data sources their businesses relied on to train or deploy AI tools.
• 37% restricted staff use of unauthorized AI tools, but 63% still operated without such restrictions.

The ISO and data-source results show that oversight remained uneven even where organizations maintained policies. A formal program or standardized assessment did not necessarily extend to worker restrictions or visibility into the data behind deployed systems.

Interest in more autonomous systems is also growing, according to BSI. About 88% of Australian businesses were exploring, piloting or deploying agentic AI.

Respondents cited security concerns, a lack of governance and unclear regulation as major obstacles. The survey therefore placed adoption and control on the same trajectory: most companies were moving toward agentic systems, but fewer than four in ten had formal governance.

Where is revenue exposed?

Alma Economics' economic model translated AI failure modes into revenue exposure. It tied the management-consulting estimate to fabricated citations, flawed datasets and hallucinated analysis.

• Global management consulting: 3.46% of sector revenue at risk, equal to USD $11.42 billion.
• Asia-Pacific management consulting: USD $3.14 billion at risk.
• Australian management consulting: AUD $169.4 million in annual revenue at risk within a sector worth AUD $4.9 billion.

The Australian estimate drew on a case in which a professional services firm produced a government-commissioned report containing fabricated citations and quotes. The errors led to revisions and a refund equal to 22.3% of the contract fee.

For autonomous and agentic AI, Alma Economics estimated that runaway computing costs, recursive logic loops and unsupervised workflows exposed 3.58% of the market each year, equal to USD $409.9 million globally.

In customer-facing aviation chatbots, booking and ticketing errors exposed USD $458.08 million of passenger air transport revenue worldwide, according to the model.

Legal services carried separate exposure of 1.28% of legal market value from non-existent case law and faulty filings, the study said.

Algorithmic bias in tenant screening and student-loan underwriting created smaller but still material exposure. The model did not assign the same revenue percentage to those use cases.

Taken together, Alma Economics' estimates give operational governance a measurable price. They also show that the exposure extends beyond generated text into autonomous workflows, customer transactions, legal filings and automated decisions.

What are business leaders calling for?

Susan Taylor Martin, Chief Executive, BSI, said:

As AI reshapes our businesses, economies and society, business leaders need to consider what happens when AI doesn't work as intended. What's clear is that AI failure is becoming a regular occurrence, bringing with it significant costs.

We need to encourage leaders to do everything they can to prevent failures while being comfortable sharing what they have learnt when an issue arises.

Realising the benefits of AI requires taking responsibility for how it is used. Business has been here before, responding to growing cybersecurity risks over the past two decades by embedding safeguards proactively and rigorously investigating and implementing learnings from AI incidents.

Now, we must prioritize putting in place robust governance and guardrails to enable safe innovation that protects us all. Acting early is likely to help reduce exposure to financial and legal liabilities, while also helping to build trust.

Charlene Loo, Managing Director, BSI Australia, said:

AI is moving quickly, and Australian businesses are moving with it. But as adoption accelerates, so does the responsibility to get it right.

When AI fails, the consequences can be significant, financially, legally and reputationally. Strong governance, independent oversight and rigorous testing need to be built in from the start. Businesses that want to benefit from AI also need to earn and maintain the trust of the people who use it.

BSI's forward-looking case is specific: early governance can reduce financial and legal exposure before agentic adoption widens failures. With 77% of Australian respondents expecting AI budgets to rise, 88% exploring, piloting or deploying agentic AI and only 37% holding formal governance programs, oversight remains less established than deployment.