Latest
Five Eyes agencies issue a call to action on AI[3]
The agencies said AI is shrinking the window between vulnerability discovery and exploitation, and asked leaders to assess risk and readiness.
Overview
AI cyber attacks: what is changing and what to do
Attackers are using AI to make scams more convincing and to find and exploit flaws faster, and AI tools are now a target in their own right. The UK's NCSC expects AI mostly to sharpen existing methods, so sound basics still do most of the work.
Last reviewed against NCSC (UK), NCSC New Zealand and UK Government (DSIT and Home Office). Sources
At a glance
- 43%
- of UK businesses identified a breach or attack in the past 12 months[5]
- 24%
- of businesses using or adopting AI had security practices for its risks[5]
- HK$200m
- sent to criminals after a deepfake video call at Arup, about £20m[7]
- #1
- prompt injection's rank in OWASP's 2025 list of LLM application risks[8]
The latest
AI cyber attacks: what has happened so far
Newest first. Each entry links to its source.
NCSC NZ publishes frontier AI vulnerability guidance[4]
The guidance says threat actors are increasingly using AI to discover and exploit vulnerabilities, and sets out steps for network defenders.
Cyber Security Breaches Survey 2025/2026 published[5]
The survey found 43% of businesses had identified a breach or attack, and that about a quarter of those using or adopting AI had security practices for it.
The full timelineIn brief
AI attacks in three answers
What the term covers, how serious it is and who should act.
What counts as an AI cyber attack
The phrase covers three different things. Attackers use AI to improve familiar attacks such as phishing and impersonation. They use it to speed up the hunt for software flaws. And they go after AI systems themselves, for example by feeding an LLM hostile instructions [2]. The NCSC's view to 2027 is that AI will almost certainly make parts of intrusion operations more effective [1].
How serious it is
The Five Eyes cyber agencies said in June 2026 that AI "is not a future consideration" and is already shrinking the window between vulnerability discovery and exploitation [3]. The NCSC adds that the time from disclosure to exploitation has already fallen to days [1]. At the same time, the NCSC expects AI mostly to enhance existing methods rather than create new threat types [1].
Who should act
Every organisation with staff who approve payments, systems on the internet, or AI tools in use. The 2025/2026 government survey found that 43% of UK businesses had identified a breach or attack over the past year [5]. Start with whether you are exposed, then how to defend.
The main routes
Where AI shows up in attacks
The sources point to four routes, each with its own page.
Impersonation and fraud
An employee at Arup sent HK$200m (about £20m) to criminals after joining deepfake video calls with impersonated senior officers [7]. See deepfake fraud.
Phishing and social engineering
The NCSC says criminals already use AI for social engineering and expects that to raise the volume and impact of intrusions [1]. See AI phishing.
Faster exploitation
NCSC guidance says threat actors are increasingly using AI to discover and exploit vulnerabilities [4]. See AI and vulnerability discovery.
Attacks on AI tools
OWASP ranks prompt injection first among its risks for LLM applications [8]. See prompt injection.
Start here
One thing to do first
First action
List every system you run that can be reached from the internet. The NCSC calls this the first step, because you cannot protect or patch what you have not found [4].
What to do
Four things to do first
-
1
List what you expose to the internet
IT and security teams
-
2
Patch frequently and promptly
IT and security teams
-
3
Verify money and access requests another way
Finance, HR and service desk teams
-
4
Limit what your AI tools can do
Developers and system owners
There are seven steps in all, with who does each one. See every step
Am I affected?
Answer a few questions
Check if you are affected
Answer three questions about your people, your internet-facing systems and your AI tools to see where to look first. Nothing you answer leaves this page.
Can anyone approve a payment, a bank detail change or an access request on the strength of a call, video meeting or message alone?
Quick answers
Questions people ask
What are AI cyber attacks?
They are attacks that use AI to improve phishing and impersonation, to find and exploit software flaws faster, or that target AI systems themselves. The NCSC expects AI to make parts of intrusion operations more effective[1]. Prompt injection is an example of an attack on AI systems[2].
Are AI cyber attacks really new?
Mostly they improve existing methods. The NCSC expects AI to increase the volume and impact of intrusions by enhancing existing techniques rather than creating novel threat vectors[1]. The Five Eyes agencies nonetheless say AI is shrinking the time between discovery and exploitation[3].
About this site
Primary sources, dated
The CyPro Consulting Ltd team maintains this site. Every version, date and figure comes from 5 named sources, including NCSC (UK), NCSC New Zealand, UK Government (DSIT and Home Office), and links to the one it came from. Last checked 4 November 2026.
Who we are and how pages are kept currentStart here
The guides, in the order to read them
Am I exposed?
Check whether your people, internet-facing systems and AI tools are exposed to AI-assisted attacks, with sourced pointers for UK organisations.
Read the guide
How to defend
A sourced plan for UK organisations to defend against AI-assisted attacks: exposure, patching, payment checks, AI tool limits and breach readiness.
Read the guide
Timeline
A dated, sourced timeline of AI-related cyber attack cases and UK and Five Eyes guidance, from the Arup deepfake fraud to the 2026 call to action.
Read the guide
Questions
Sourced answers to common questions on AI cyber attacks: deepfakes, phishing, faster exploitation, prompt injection and what UK organisations should do.
Read the guideGo deeper
In depth Deepfake fraud 
In depth AI phishing 
In depth Faster exploitation 
In depth Prompt injection 
Question What they are 
Question Are they new? 
In depth Your AI tools 
In depth Board briefing 
Question UK statistics How CyPro can help
Need a hand with any of this?
Patching is one job. Keeping a vulnerable system safe until it is patched, and proving it, is several more. CyPro's consultants, testers and SOC analysts can take any of them off your plate.
Vulnerability scanning
Find every internet-facing system you run, including the forgotten ones, and see which are still on a vulnerable build.
See the scanning service
Hardening and segmentation
Our consultants put the controls in place that limit the damage when something is missed: firewall rules, network isolation and tighter admin access.
More on cypro.co.uk
24/7 security monitoring
Our SOC analysts watch your systems day and night for the signs of an attack in progress: unexpected processes, new outbound connections, logins that do not fit.
More on cypro.co.uk
Penetration testing
A focused test of your systems and the controls around them, so you know they hold before an attacker finds out for you.
More on cypro.co.uk
Incident response
If you may already be compromised, our responders contain it, preserve the evidence, find out what was touched and get you running again safely.
More on cypro.co.uk
Fractional CISO for the board
A senior security lead who advises the board on what to fix first, what to accept and what to tell customers, and owns the decision record.
More on cypro.co.uk
Know your exposure
Not sure you are protected?
AI makes it quicker to find and use weak points, so it helps to find them first. A penetration test shows which of your internet-facing systems, staff processes and AI tools an attacker could actually use.