Something happened this week that security experts have been warning about for years. And now that it has happened once, the world of cybersecurity will never be the same.
For the first time in documented history, hackers used artificial intelligence to develop a zero-day exploit a cyberattack that targets a vulnerability nobody knew existed and deployed it in a real, live attack. Not in a lab. Not in a simulation. In the real world, against real targets.
At the same time, a sophisticated supply chain attack hit multiple widely-used software platforms simultaneously including tools tied to Mistral AI, UiPath, and TanStack. And a separate breach exposed student and staff records at some of the world’s most prestigious institutions, including MIT and Oxford.
This wasn’t a coincidence. This was a week that changed the cyber threat landscape permanently.
What exactly is a zero-day exploit — and why is this so serious?
A zero-day exploit is an attack that takes advantage of a software vulnerability that the software’s developers don’t yet know about. The name comes from the fact that developers have had “zero days” to fix it. There is no patch. There is no defence. By the time anyone knows the gap exists, it has already been used against them.
Zero-day exploits have always been the most feared weapon in a hacker’s arsenal. They are rare, expensive, and until now, required highly skilled human researchers to discover. Governments and intelligence agencies have paid millions of dollars for a single zero-day vulnerability.
What happened this week changed that equation entirely. AI found one. Automatically. And then it was used in an attack.
“AI is having a seismic impact on cybersecurity — but so far that impact has been mostly one-sided. Cybercriminals are happily taking advantage of AI agents to increase the speed, frequency and sophistication of attacks.” — GeekWire, May 11, 2026
Why should YOUR business care about this?
You might be reading this thinking: “this sounds like a problem for big tech companies and universities. Not for my business.” That thinking is exactly what cybercriminals count on.
Here is the reality. AI-powered attack tools do not discriminate by company size. They scan everything. They probe everything. And when they find a gap in a small business website, an outdated plugin, a weak employee password they exploit it with the same speed and efficiency they would use against a multinational corporation.
The difference is that large companies have dedicated security teams. You probably don’t. That makes you easier, not harder, to attack.
What every business must do right now.
The good news is that basic cybersecurity hygiene blocks the vast majority of attacks, even AI-powered ones. Here is where to start:
-
Update everything immediately. The supply chain attack this week exploited software tools thatbusinesses trusted. Check every tool, plugin, and platform you use and ensure it’s running the latest version.
-
Enable two-factor authentication on every account. This one step blocks over 99% of automatedcredential attacks, including those using stolen AI tool credentials.
-
Audit who has access to what. Many breaches succeed because too many people have access totoo many systems. Limit access to only what each person genuinely needs.
-
Back up your data today and test that the backup works. A working backup is your last line ofdefence against ransomware.
-
Work with developers who build security in. If your website or software was not built with securityas a priority, it is a liability, not an asset.
The three attacks that defined this week.
The AI Zero-Day Attack
AI was used to identify and exploit a previously unknown software vulnerability in a live attack. This is a historic first. It means the time between a vulnerability existing and being exploited has collapsed from months to potentially hours. Human defenders simply cannot keep pace with AI-driven discovery at this speed.
The Canvas Data Breach
Hackers breached Canvas school login portals, exposing student and staff data at MIT, Oxford, and other major universities. The breach is a reminder that no organisation — regardless of size, budget, or prestige — is immune. If the data exists and the door is weak, someone will eventually walk through it.
The Supply Chain Attack
A coordinated attack compromised npm packages tied to Mistral AI, UiPath, and TanStack — three platforms used by thousands of developers and businesses worldwide. Supply chain attacks are particularly dangerous because they don’t target you directly. They poison the tools you trust, reaching you through software you already have installed.