PUBLISHED DATE: 20 November 2026
AI security tools originally developed to help defenders, such as ARTEX, are being repurposed by cybercriminals, accelerating the pace and scale of cyber-attacks, while reducing the expertise needed to enact them.
These AI tools can autonomously identify and exploit vulnerabilities, steal credentials (usernames and passwords), find weaknesses in your businesses systems and networks, and move through networks after gaining access. This automation allows cybercriminals to undertake multiple attacks simultaneously, operating at a far greater speed and scale than ever before, and targeting more businesses and organisations increasing their chances of success.
The result is a growing risk of more frequent and disruptive cyber incidents. Activities that once took days or weeks now occur at machine speed, often faster than businesses and organisations can detect and respond. This is how AI is supercharging cyber risk, increasing both the likelihood of cyber threats and the potential impact on business operations, services, and resilience.
The NCSC is urging New Zealand businesses and organisations to be proactive in protecting internet-connected tools and systems. No business or organisation is immune to the capabilities of these AI-enabled cyber threats, and it is more important than ever to put the right protections in place and adopt a mindset that a cyber incident could happen at any time.
Practical protections to reduce cyber risk
While it is impossible to stop cyber attacks altogether, the impact and effect of AI-enabled attacks can be reduced through understanding your businesses cyber risk and what protections are required to best stop or mitigate an attack. This can be best achieved through “defence-in-depth" - when you implement multiple different security measures to ensure that if one fails, another one can help prevent and mitigate security risks.
1. Implement multi-factor authentication (MFA)
Many AI-enabled attack tools obtain initial access through compromised credentials.
Implementing MFA ensures that if your business's credentials do become compromised, there are additional measures that can stop the attack.
2. Reduce the opportunities for attack by limiting internet-connected devices and systems
Exposing more devices and systems to the internet increases the areas of opportunity for an attack to get in and obtain access. Ensure that anything that doesn’t need to be connected to the internet isn’t. Especially if you have old, unused or out-of-date software or devices.
Ensure that remote access to the network is removed or highly restricted and monitored.
3. Secure your network
Your business network is all the connected devices, applications and accounts used within your business. If an AI-enabled attack tool gains access to your systems, it will attempt to move throughout the network to obtain additional credentials, data, or obtain higher privileges to gain further access to your systems and data.
Talk to an IT or network engineer to explain what your business does, and what you use your business network for. They can help you configure any separate networks or network devices that you may need to protect yourself.
4. Restrict employee access to networks and devices
AI-enabled attack tools will take advantage of accounts and systems where additional, un-necessary access is provided. Consider restricting access to parts of your network, giving people the lowest level of access they need to do their jobs.
This is the 'principle of least privilege'. You can set different permissions for each role based on the level of access they need. Staff should only have access to the systems, files and information they need to do their job.
5. Keep devices and systems up-to-date and plan to remove out-of-date technology
AI-enabled attack tools can rapidly find and exploit, old, out-of-date technologies that are no longer being updated, patched or maintained.
Ensuring software is up to date removes the likelihood that older, known vulnerabilities are exploited. Furthermore, decommissioning old services, technologies and accounts ensure that they are not forgotten and left unmaintained.
6. Set up logs and monitoring
Attacks using AI-enabled tools are typically noisy and should be noticeable through standard monitoring tools.
Keeping logs centralised in one place and monitoring them can alert you to an incident early, so you can stop it before it goes too far. Logs are also important for understanding how an incident occurred, when it started, and which systems or accounts were affected. Knowing when an action happened and who took it means you can resolve incidents quicker and get back to business as usual.
7. Prepare for rapid incident response
If your business has a cyber security incident, you’ll need to know what steps to take to keep your business running. Having a clear plan in place will help you through what could be a stressful time. It’ll help your team respond to an incident quickly and improve your business's resilience.