Attackers are now using AI to break into websites and business systems with nobody at the keyboard. The FBI says some AI-driven attacks went from choosing a target to high-level access to its systems in just a few hours.
I’ve followed AI closely for the past 24 months, and even more closely over the last two or three weeks with everything that’s been going on. I’ve spent over 25 years working across marketing, business and IT, with experience spanning cybersecurity, governance and risk. AI is now central to how we operate at Outsource2Us, and in all my years working in technology, I’ve never seen the pace of change we’re experiencing today.
Then it hit a little closer to home. One of our own clients was recently compromised.
I’m not going to share who they are, what industry they’re in or how it happened. But here’s what concerns me. I don’t believe this will be an isolated incident, and I certainly don’t see things slowing down anytime soon.
So, if you own a business, I’d encourage you to read on.
I’m not here to scare anyone or add to the hype. I want to share what’s actually been happening over the past two weeks, show you the sources so you can check the facts yourself, and give you eight practical things you can do this week to better protect your business.
What has happened in the past two weeks
There’s a lot of talk about AI and cyberattacks at the moment, and with so much information (and misinformation) circulating online, it’s becoming increasingly difficult to know what to believe or who to trust.
That’s one of the reasons I’m putting my hand up to help, I think it’s important to share what I’m seeing, particularly when it could affect so many businesses.
So rather than add to the noise or unnecessary fear, let’s look at what’s actually happening, what the experts are reporting, and what it means for you and your business.
The FBI has warned about a new AI-powered hacking tool
On 13 November the FBI put out a warning about a tool called ARTEX.
And this is where things get interesting, and frankly, pretty concerning.
ARTEX is publicly available online and can connect to AI models similar to those powering the chatbots millions of people use every day.
The difference is what it’s being used for.
Instead of answering questions or helping someone write an email, it can automate much of the process of breaking into a business’s systems, from identifying vulnerabilities and gaining access through to stealing sensitive information. And it requires surprisingly little human involvement.
According to the FBI, attackers have been using the tool since at least August, targeting businesses and organisations across financial services, healthcare, government, legal services and other industries.
In some cases, attackers went from identifying a target, to gaining high-level access to its systems in less than three hours.
Three hours. Think about that.
But here’s the part that really caught my attention. The FBI warns that people with very little technical knowledge can now use tools like this to carry out sophisticated cyberattacks against multiple targets at once.
One campaign targeting South Korean financial institutions reportedly compromised the personal information of approximately 67,000 people. And that’s what I think every business owner needs to understand…
…Not long ago, carrying out this level of attack required serious technical expertise. AI is now putting those capabilities into the hands of people who previously wouldn’t have had the skills to pull it off.
Google has found AI carrying out cyberattacks on its own
Just four days later, on 17 November, Google’s threat intelligence team reported something equally concerning.
Over the previous three months, it had identified attackers using AI to find weaknesses in software and exploit them, without a human needing to control the process.
One of the tools Google discovered was particularly interesting.
It monitors popular website software for security updates. When a new update is released, the AI compares the old version with the new one, identifies the security weakness that was fixed, and then creates and launches an attack against websites that haven’t updated yet.
All automatically.
Think about that for a moment. An attacker doesn’t even need to be sitting at a computer.
Now, to be fair, Google also pointed out that these tools aren’t perfect. They make mistakes, they’re often easy to detect, and not every AI-generated attack actually works.
But that’s not really the point.
AI doesn’t need to be perfect to break into an outdated website. It just needs to find the weakness before someone fixes it.
Banks and government agencies have also been making headlines
During the same week, JPMorgan Chase, one of the world’s largest banks, advised customers that some of its external service providers had been targeted in a cyberattack.
The bank said customer information, including names and email addresses, may have been exposed.
Then, on 22 November, the ABC reported that the FBI was investigating a possible breach of its own national crime database.
At the time of reporting, the FBI hadn’t publicly confirmed the incident, and there was no confirmed information about who might have been responsible.
So that one is still developing.
It’s already happening here in Australia
And this isn’t just an overseas problem.
In June, an AI agent belonging to OpenAI, the company behind ChatGPT, accessed Medicare statistics that weren’t publicly available through a Services Australia online portal.
An AI agent is essentially AI that can take actions on its own, rather than just answering questions or following instructions one step at a time.
What’s concerning about this incident is that nobody specifically instructed the AI to access that information.
It did it independently.
The Prime Minister confirmed that no personal information was accessed.
However, OpenAI reportedly didn’t discover what had happened until August, and the Australian Government wasn’t notified until September.
On 10 November, the government accepted all 18 recommendations from its review and announced a new AI safety regulator.
So we’re talking about an AI system accessing information it shouldn’t have, without anyone specifically telling it to, and the incident not being identified or reported for months.
That’s a pretty clear example of why AI security is becoming such an important issue.
Why small business owners should be paying attention
Now, you might be thinking, “That’s all interesting, Damo, but we’re talking about banks, government departments and the FBI. Why would anyone want to hack my little business website?”
It’s a fair question. And I think the answer is actually pretty simple…
Nobody needs to be specifically targeting your business anymore. Attackers can use automated tools to scan large numbers of websites, looking for weaknesses they can exploit.
Your business doesn’t need to be famous, successful or even particularly interesting.
If your website has a security weakness, it can become a target simply because the software found it.
And we already know this is affecting Australian businesses.
The Australian Signals Directorate, which oversees the Australian Cyber Security Centre, issued an alert back in July about a large-scale campaign targeting website software.
According to the alert, many small and medium-sized Australian businesses have already been affected.
A large number of the identified weaknesses involved WordPress plugins.
For anyone unfamiliar with WordPress, it’s one of the most widely used platforms for building websites. Plugins are the additional tools that provide features such as contact forms, booking systems, security and backups.
Here’s the concerning part.
Many of the security weaknesses being exploited were already known, and updates were available to fix them.
In other words, some of these attacks could potentially have been prevented by keeping the website software up to date.
And AI is making the situation more urgent.
The Australian Cyber Security Centre has warned that attackers can use AI to identify weaknesses, exploit them quickly and target large numbers of websites at once.
On 17 November, it issued another warning encouraging organisations to install security updates faster.
Some weaknesses are now being exploited within just 48 hours of a fix becoming available.
The agency’s latest Annual Cyber Threat Report summed up the changing situation:
“In the past, defenders had months to patch, now they have days, sometimes only hours.”
Australian Signals Directorate, Annual Cyber Threat Report 2025-26
And the financial consequences aren’t small either.
The same report found that small businesses reporting cybercrime experienced an average self-reported cost of $55,400.
That’s a serious amount of money for any small business.
The message here is pretty straightforward.
The days of leaving website updates for a few weeks or months and hoping everything will be fine are coming to an end.
The rules are changing for all of us
I’ll be upfront about something…Website developers, hosting companies, IT providers and marketing agencies are all having to adapt to these changes.
And that includes us at Outsource2Us.
We’ve been building, managing and securing websites for years, but the speed at which these threats are developing is something different.
I’d be very cautious of anyone who tells you their website or systems are completely secure and can’t be hacked.
That’s simply not a promise anyone can realistically make in this current climate. And you don’t need to take my word for it.
In June, the heads of the cybersecurity agencies from Australia, the United States, the United Kingdom, Canada and New Zealand issued a joint statement to business leaders.
These are the five countries that make up the Five Eyes intelligence partnership.
They warned that the latest AI systems are changing cybersecurity on both sides, helping attackers become more capable while also giving defenders better tools.
One sentence really stood out to me:
“The timeline is not years, it is months.”
Five Eyes cybersecurity agencies, joint statement, June 2026
And further into the statement, they made another point:
“Breaches will occur.”
When the cybersecurity agencies responsible for protecting five countries are making statements like that, I think every business owner should be paying attention.
But there’s some good news here too.
The same AI technology being used by attackers is also helping businesses protect themselves.
AI can help security teams identify weaknesses earlier, monitor systems, detect unusual activity and respond faster when something goes wrong.
And I expect those capabilities will continue to improve.
AI is already a central part of how we operate at Outsource2Us, so we’re seeing firsthand just how quickly this technology is advancing.
It’s one of the reasons I wanted to write this article. I’d much rather help business owners understand what’s happening now than see more people learn about it after their website or business has been compromised.
The bigger picture: AI is moving faster than most people realise
I genuinely think 99% of people have very little idea just how quickly AI is developing, and what the frontier labs are actually doing.
Most people are busy running businesses, looking after staff, dealing with customers and getting on with their lives.
They don’t have the time to spend hours every day following what’s happening in AI.
But I think we’re approaching a period where understanding this technology is going to become increasingly important, whether you’re interested in AI or not.
If you’re curious about where all of this might be heading, there are three topics I’d encourage you to explore.
1. AI 2027
AI 2027 is a detailed scenario published in April 2025 by a nonprofit research organisation called the AI Futures Project.
Its lead author, Daniel Kokotajlo, previously worked as a researcher at OpenAI before leaving in 2024 and calling for greater transparency in the AI industry.
The report explores what could happen if AI becomes capable of carrying out advanced AI research itself.
In simple terms, what happens when AI starts helping to develop the next generation of AI?
The researchers originally identified 2027 as a likely timeframe for some major developments.
They revised their estimates during 2025, and on 10 November this year brought their predictions forward again.
Their latest preliminary assessment suggests that AI capable of performing research at the level of human AI researchers could arrive sometime in 2027, unless development is deliberately slowed.
Now, these are predictions, not facts.
The authors themselves acknowledge the uncertainty, and nobody knows exactly when these capabilities might arrive.
But I think the research is worth reading, even if you don’t agree with the predictions.
2. Recursive self-improvement
This is one of the more interesting concepts in AI.
Recursive self-improvement is the idea that AI could become capable of improving its own intelligence and abilities.
Then, the improved version could develop an even better version of itself.
And that version could do the same thing again.
You can probably see where this starts getting interesting.
What’s remarkable is that the Australian Government’s own cyber threat reporting now identifies recursively self-improving AI as a potential factor that could accelerate future cybersecurity threats.
This isn’t proof that such systems already exist or that uncontrolled improvement is inevitable.
But it’s significant enough that government cybersecurity agencies are now seriously considering the implications.
3. The technological singularity
This is probably the biggest concept of the three.
The technological singularity describes a hypothetical point where artificial intelligence becomes more capable than human intelligence, potentially leading to technological progress so rapid that we can no longer reliably predict what comes next.
Now, nobody knows whether this will happen, when it might happen or exactly what it would mean.
There are some very smart people who believe it could happen relatively soon, and others who think we’re still a long way off.
Personally, what I find most interesting is that these conversations are now happening seriously among researchers and people building the technology.
Meanwhile, most of the business community is barely aware the conversation is taking place.
I think that’s a pretty significant gap.
I’m not suggesting everyone needs to become an AI expert.
But I do think business owners should start paying more attention to what’s coming.
The implications for business, employment, productivity, security and everyday life could be enormous.
And that’s probably a conversation for another article.
Eight things every business owner should do this week
So, after all of that, what can you actually do to protect your business?
The good news is that you don’t necessarily need to spend thousands of dollars or become an IT expert.
There are some simple, practical steps you can take right now.
Most of these recommendations come directly from Australian Government cybersecurity guidance.
1. Find out who’s responsible for your website security
This sounds obvious, but you’d be surprised how many business owners don’t actually know.
Is it your web developer? Your hosting company? Your IT provider? Your marketing agency?
Don’t assume someone is looking after it.
Find out exactly who is responsible for keeping your website secure and up to date.
You should have a clear answer and ideally one person you can contact when something goes wrong.
2. Ask your website provider four simple questions
You don’t need to understand the technical details. You just need to know someone is managing them properly.
Ask whoever looks after your website:
- Is all the software on my website still supported and receiving security updates?
- Are security updates installed automatically or as soon as reasonably possible?
- Is the software we’re using reputable and configured securely?
- If our website was hacked, how quickly would you know, notify me and get it back online?
If they can’t answer these questions clearly, I’d be concerned.
3. Make sure security updates are being installed
Website software needs regular updates, just like your computer and phone.
These updates don’t just add new features. They often fix security weaknesses that attackers are actively looking for.
Ask your website provider whether automatic updates are enabled where appropriate, and how they handle more significant updates that need testing.
And while you’re at it, make sure your computers and phones are being updated too.
4. Get rid of software and accounts you no longer use
Old website plugins, outdated themes, unused user accounts and forgotten websites can all introduce unnecessary security risks.
If you’re not using something, ask whether it can be safely removed.
And if you’ve got old websites sitting online that nobody looks after anymore, it’s worth reviewing those as well.
The fewer unnecessary things connected to your systems, the fewer potential ways an attacker has to get in.
5. Turn on multi-factor authentication
Multi-factor authentication, often called MFA or two-factor authentication, adds another layer of protection to your accounts.
Instead of relying on just a password, it requires an additional verification step, usually through an app or your phone.
This makes it much harder for someone to access your accounts using a stolen password.
At a minimum, I’d recommend enabling it on your website administration account, web hosting, domain registration and business email.
6. Check your website backups
If your website was compromised tomorrow, could you get it back online?
And how long would it take?
Backups are an essential part of protecting your business, but simply having them isn’t enough.
You need to know they’re working.
Ask your website provider when the last backup was taken, where it’s stored and whether they’ve tested restoring the website from it.
A backup that doesn’t work when you need it isn’t much use.
7. Ask your provider to check the latest government security alerts
The Australian Cyber Security Centre regularly publishes alerts about known security issues affecting businesses.
Forward the latest website security alert to whoever manages your website and ask them to check whether any of the affected software is being used.
If it is, ask them to confirm that the necessary updates or protections have been applied.
And get that confirmation in writing.
8. Know who to contact if something goes wrong
If you suspect your website or business systems have been compromised, don’t ignore it.
Contact your website or IT provider immediately.
You can also report cybersecurity incidents to the Australian Cyber Security Centre at cyber.gov.au/report.
Or call the Australian Cyber Security Hotline on 1300 CYBER1 (1300 292 371).
The hotline operates 24 hours a day.
It’s worth saving those details somewhere you can access even if your business systems are unavailable.
If you want to take things further, the Australian Signals Directorate also provides its Essential Eight cybersecurity recommendations, along with its newer ASD Essentials framework.
You can find practical advice for Australian small businesses through the Small Business Hub at cyber.gov.au.
But my advice is simple.
Start with the first two steps this week. Find out who’s responsible for your website security and ask those four questions. One phone call or email could tell you a lot about how well your business is protected.
A final word from me
Look, I understand it’s easy for me to talk about website security.
We build and manage websites for businesses every day at Outsource2Us, so naturally this is something we take seriously.
But I also believe that gives us a responsibility to help business owners understand the risks, particularly when the technology is changing this quickly. And I’ll be the first to say that nobody can guarantee a website will never be hacked. Not us, not your hosting provider, not your IT company, and certainly not anyone selling you a supposedly unbreakable security solution.
What we can do is take sensible precautions, stay on top of updates, monitor our systems and make it as difficult as possible for attackers to get in.
If you’re unsure about the security of your website, feel free to reach out to us.
We’re happy to take a look and give you a straightforward assessment of anything we think needs attention.
I’m not sharing all of this to frighten anyone. I’m sharing it because I’ve spent more than 25 years working across business, marketing and IT, and I’ve never seen technology change at anything approaching this pace.
And I think the next few years are going to be even more interesting.
So I’ll leave you with one question.
When did you last ask who is looking after the security of your business website?
– Damo