Microsoft Copilot could have been hacked by some very low-tech methods

Copilot imagery from Microsoft
(Image credit: Microsoft)

Cybersecurity researchers have found a way to force Microsoft 365 Copilot to harvest sensitive data such as passwords, and send them to malicious third parties using “ASCII smuggling”

The ASCII smuggling attack required three things: Copilot for Microsoft 365 reading the contents of an email, or an attached document; having access to additional programs, such as Slack; and being able to “smuggle” the prompt with “special Unicode characters that mirror ASCII but are actually not visible in the user interface.”

As the researchers at Embrace the Red, who found the flaw, explain, Microsoft 365 Copilot can be told to read, and analyze, the contents of incoming email messages and attachments. If that email, or attachment, tells Microsoft 365 Copilot to look for passwords, email addresses, or other sensitive data in Slack, or elsewhere, it will do as it’s told.

Hidden prompts and invisible texts

Ultimately, if such a malicious prompt is hidden in an attachment, or email, via special Unicode characters that render it invisible to the victim, they may end up, unknowingly, telling their AI chatbot to hand over sensitive data to malicious third parties.

To prove their point, the researchers shared exploit demos with Microsoft, showcasing how sensitive data, such as sales number and multi-factor authentication (MFA) codes, can be exfiltrated and then decoded.

“An email is not the only delivery method for such an exploit. Force sharing documents or RAG retrieval can similarly be used as prompt injection angles,” the report concludes.

In the paper, the researchers recommended Copilot 365 stops interpreting, or rendering, Unicode Tags Code Points.

“Rendering of clickable hyperlinks will enable phishing and scamming (as well as data exfil),” the report concludes. “Automatic Tool Invocation is problematic as long as there are no fixes for prompt injection as an adversary can invoke tools that way and (1) bring sensitive information into the prompt context and (2) probably also invoke actions.”

Microsoft has since addressed the issue.

More from TechRadar Pro

Sead is a seasoned freelance journalist based in Sarajevo, Bosnia and Herzegovina. He writes about IT (cloud, IoT, 5G, VPN) and cybersecurity (ransomware, data breaches, laws and regulations). In his career, spanning more than a decade, he’s written for numerous media outlets, including Al Jazeera Balkans. He’s also held several modules on content writing for Represent Communications.

Read more
hacker.jpeg
Thousands of GitHub repositories exposed via Microsoft Copilot
A concept image of someone typing on a computer. A red flashing danger sign is above the keyboard and nymbers and symbols also in glowing red surround it.
Microsoft Teams and other Windows tools hijacked to hack corporate networks
Microsoft
"Another pair of eyes" - Microsoft launches all-new Security Copilot Agents to give security teams the upper hand
A fish hook is lying across a computer keyboard, representing a phishing attack on a computer system
Microsoft authentication system spoofed via phishing attack
Copilot on a laptop
Microsoft quietly updates Copilot to cut down on unauthorized Windows activations
A fish hook is lying across a computer keyboard, representing a phishing attack on a computer system
Microsoft 365 accounts are under attack from new malware spoofing popular work apps
Latest in Security
cybersecurity
Chinese government hackers allegedly spent years undetected in foreign phone networks
Data leak
A major Keenetic router data leak could put a million households at risk
Code Skull
Interpol operation arrests 300 suspects linked to African cybercrime rings
Insecure network with several red platforms connected through glowing data lines and a black hat hacker symbol
Multiple routers hit by new critical severity remote command injection vulnerability, with no fix in sight
Code Skull
This dangerous new ransomware is hitting Windows, ARM, ESXi systems
An abstract image of a lock against a digital background, denoting cybersecurity.
Critical security flaw in Next.js could spell big trouble for JavaScript users
Latest in News
DeepSeek
Deepseek’s new AI is smarter, faster, cheaper, and a real rival to OpenAI's models
Open AI
OpenAI unveiled image generation for 4o – here's everything you need to know about the ChatGPT upgrade
Apple WWDC 2025 announced
Apple just announced WWDC 2025 starts on June 9, and we'll all be watching the opening event
Hornet swings their weapon in mid air
Hollow Knight: Silksong gets new Steam metadata changes, convincing everyone and their mother that the game is finally releasing this year
OpenAI logo
OpenAI just launched a free ChatGPT bible that will help you master the AI chatbot and Sora
An aerial view of an Instavolt Superhub for charging electric vehicles
Forget gas stations – EV charging Superhubs are using solar power to solve the most annoying thing about electric motoring