Keyboard with a button to generate a report

What ASD’s 2025-26 Threat Report Means for Mid-Sized Businesses

In 2025-26, 362 cyber security incidents were only identified after the Australian Signals Directorate alerted the organisation to potential malicious activity on its network.

That figure sits deep in ASD’s Annual Cyber Threat Report 2025-26, well behind the headlines about 1,500 incidents and a 20% rise. It’s also the more telling one. Almost one in four incidents ASD responded to came to light after ASD raised the alarm.

ASD counts unsuccessful attempts as incidents too, so not every one of these was a breach. But in each case, ASD flagged activity on the network before the organisation had identified it.

The report’s story for mid-sized businesses comes down to cost and visibility. Medium business losses have risen two years running, and in 2025-26, a significant share of incidents only surfaced after a warning from ASD.

Key findings for mid-sized businesses

Medium business costs have risen two years running

The average self-reported cost of cybercrime for a medium business was $62,800 in 2023-24, $97,200 in 2024-25 and $141,300 in 2025-26. That’s an increase in each year, and more than double over two.

The other business sizes haven’t followed a consistent pattern. Small business costs barely moved over the same period, and sit at $55,400. Large business costs rose 219% and then fell 27%, to $148,100.

These are self-reported averages from ReportCyber, so any single year deserves caution. What stands out for medium businesses is that the increase has held two years in a row.

Ask your team: if operations were disrupted for a week, what would it cost the business, and has anyone tested that estimate?

Incidents are going deeper

ASD rates each incident on two things: the damage it causes, and how significant the affected organisation is. In 2025-26, fewer incidents reached its highest severity categories, while extensive compromises (i.e. incidents with an extensive impact, the second-worst of ASD’s six impact levels) rose from 26 to 40. Ransomware incidents held steady at 139, while extensive ransomware incidents rose from 16 to 24. ASD also updated how it categorises incidents, but these comparisons are its own.

The further an attacker gets before anyone notices, the more systems need investigating, restoring and proving clean.

Manufacturing deserves particular attention. It was one of the three sectors most affected by ransomware, alongside health care and professional, scientific and technical services. It also recorded the largest increase in incident reports of any sector.

Ask your team: if an attacker had been inside the network for a week, how would we know which systems they’d reached?

The access attackers use is the access your staff use

ASD describes stolen usernames, passwords, tokens and sessions as now among the most common pathways into Australian networks. Its analysis of the techniques used in incidents puts phishing at 39%, compromised accounts at 21% and use of valid accounts at 15%.

ASD notes that attackers using valid accounts can access networks and systems without triggering traditional security controls. Signed in with a legitimate account, an attacker can be hard to tell apart from an employee. It’s one reason suspicious activity can go unnoticed until someone outside the organisation flags it.

Ask your team: could we say today which accounts have privileged access, and when each was last used?

Evidence is what changes the outcome

One case study in the report shows the alternative. In mid-2025, an Australian organisation was hit by ransomware that disrupted its security cameras, building access, lighting, lifts and more than 30 virtual servers.

Its event logs showed how the attacker got in, which systems were affected and what the attacker did. They also showed no data had been stolen. With backups in place, essential services were restored promptly.

The incident was still disruptive. But the organisation knew its scope, because it had evidence instead of assumptions.

Ask your team: if we were hit tomorrow, would our logs be complete enough, and kept long enough, to show how the attacker got in and what they touched?

The question worth asking

ASD’s advice for medium-sized businesses defending against AI-enabled threats starts with knowing your environment. Keep an inventory of internet-facing systems, patch promptly, segment your network, and test your incident response and business continuity plans.

None of that is new. What the threat report adds is a clear picture of what’s at stake for businesses your size. So the useful question isn’t whether cyber threats are increasing. It’s this: if there was suspicious activity on your network today, would your team see it before ASD did?

For what this means for your Essential Eight evidence as ASD Essentials arrives, read Is Your Security Evidence Current Enough for ASD Essentials?

FAQs

What is the ASD Annual Cyber Threat Report 2025-26, and what period does it cover?

It’s the seventh annual threat report from the Australian Signals Directorate’s Australian Cyber Security Centre (ASD’s ACSC). It covers the 2025-26 financial year, from 1 July 2025 to 30 June 2026. It examines Australia’s cyber threat environment, the impact of AI on that threat, and how organisations can protect themselves. Its statistics come from two datasets: cybercrime reported through ReportCyber, and the incidents ASD’s ACSC responded to.

How much does cybercrime cost a medium-sized business in Australia?

In 2025-26, the average self-reported cost of cybercrime for a medium business was $141,300, up 45% on the previous year. Small businesses reported an average of $55,400 and large businesses $148,100. These figures are self-reported through ReportCyber, and ASD notes that most cybercrime likely goes unreported.

How many cyber security incidents did ASD respond to in 2025-26?

More than 1,500, an increase of 20% on the previous year and 36% over five years. They included 139 ransomware incidents and 98 data breaches. ASD also notified organisations of potential malicious activity more than 4,400 times. ASD attributes part of the rise to improved cyber security awareness and reporting.

ASD’s incident figures include unsuccessful attempts, not only confirmed compromises.

What are the most common ways attackers get into Australian networks?

Phishing was the most common initial access technique ASD observed, at 39% in its analysis of incident techniques. Other common entry points were the use of valid accounts (15%) and exploiting public-facing applications (13%). ASD describes stolen usernames, passwords, tokens and sessions as now among the most common pathways into Australian networks.

Let's see how we can personalise your IT

Evolution Systems is ISO 27001 Certified

Information Security ISO 27001 Certification