ASD Threat Report: Medium Business Cyber Losses Jump As AI Speeds Up Attacks
On 26 October 2026, the Australian Signals Directorate published its Annual Cyber Threat Report 2025-26. The average cybercrime cost reported by medium businesses rose 45%, and ASD says AI is cutting the time a business has to patch before attackers move in.
On 26 October 2026, the Australian Signals Directorate (ASD) published its Annual Cyber Threat Report for the 2025-26 financial year. Its Australian Cyber Security Centre responded to more than 1,500 cyber security incidents, up 20% on the year before. It notified organisations of possible malicious activity more than 4,400 times, up almost 80%. It also received over 82,000 cybercrime reports, about one every six minutes.
The business numbers are the part owners should read closely. The average self-reported cost per cybercrime report from a business rose 16% to $93,500. For medium businesses it rose 45% to $141,300, the biggest jump of any size group. Small businesses held steady at $55,400, and large businesses fell 27% to $148,100.
What The Report Says About AI
AI runs through the whole report. The foreword, from Defence Minister Richard Marles, says AI was the main force shaping cyber threats over the past year, helping attackers find weaknesses faster and exploit them at greater scale. The executive summary is more measured. It says AI hasn’t created new kinds of threat, but it has amplified the ones that already existed.
Patching is where that shows up first. ASD has previously seen most exploitation happen within two weeks, with some flaws exploited within 48 hours of a fix or advice being released. It expects AI tools to shrink that window further over the next 12 months. In ASD’s framing, defenders used to have months to patch and now have days, sometimes hours. The report also records an attacker successfully exploiting a vulnerability first identified in 2008, so old gaps still get used too.
Stolen logins are the other big theme. The report lists stolen usernames, passwords, tokens and sessions among the most common ways into Australian networks. Compromised accounts and credentials are now one of the top incident categories across government, industry and critical infrastructure.
In an exclusive ABC interview published the same day, ASD director-general Abigail Bradshaw said: “What is inevitable is that with each capability increase, we are likely to see more disruption.” She told the ABC that disruption could mean websites taken offline to be patched, data breaches or services going down.
What It Means For An Australian Business Owner
The report gives specific advice by business size. To defend against AI-enabled threats, ASD recommends four things for medium-sized businesses: keep an inventory of your internet-facing systems, patch vulnerabilities promptly, segment your network, and test your incident response and business continuity plans. For small businesses, it suggests choosing products that are secure out of the box, turning on automatic updates and being clear about who is responsible for security monitoring and incident response.
Every organisation gets the same baseline. Turn on multi-factor authentication (MFA) or passkeys. Only expose the systems and services you actually need to the internet. ASD also says your suppliers’ cyber risks are your risks, so the software and service providers that touch your data are part of the picture.
ASD wants organisations using AI, including to strengthen their own defences, but with controls in place: give AI tools only the access they need, limit the data they can see, check their outputs before acting on them, and keep a record of what they did.
Here’s our read. Every AI tool or automation a business connects to its email, accounting software or CRM is another system with access to that data. It belongs on the same inventory as everything else, with the same questions about logins, updates and who supplies it.
The Step To Take First
Write a one-page list of every system that holds your business data or can be reached from the internet. Include email, accounting, CRM, your website, booking tools, remote access, and any AI tool or automation connected to them. Next to each one, note whether MFA or passkeys are switched on, who applies updates, and which supplier runs it.
Then turn on MFA or passkeys wherever they’re off, and switch on automatic updates wherever the product allows it. That list covers the inventory of internet-facing systems ASD asks medium businesses to keep. It also shows you which suppliers hold your data.
If something does go wrong, ASD asks businesses to report it at cyber.gov.au/report or on 1300 CYBER1 (1300 292 371).
If AI tools are on your list and you want to check how they’re connected, our guide to whether it’s safe to connect AI automation to your business data covers the controls we build in.
Sources
Go Deeper
- Is It Safe To Connect AI Automation To Your Business Data? →
How AI automation keeps your business data safe: built in your own stack, least-privilege access, human-in-the-loop, and what to ask before connecting anything.
Sam co-founded Echelon AI Solutions and leads transformation strategy, client engagements and growth. He has built and operated businesses across marketing and AI education, and has guided companies in retail, trades, hospitality and professional services through operational change. His focus is making AI earn its place through measurable business performance.