typo

Author: wunderwuzzi23

content/posts/2025/spaiware-and-chatgpt-command-and-control-via-prompt-injection-zombai.md

@@ -35,7 +35,7 @@ The prerequisite is that the user gets infected once with prompt injection malwa
 
 [![initial infection](/blog/images/2024/chatgpt-c2-initial-infection.png)](/blog/images/2024/chatgpt-c2-initial-infection.png)
 
-The initial infection persists malicious instructions in ChatGPT's long term storage.
+The initial infection persists malicious instructions in ChatGPT's long-term storage.
 
 ### Hey ChatGPT - Check-out this GitHub Issue!
 

docs/posts/2025/spaiware-and-chatgpt-command-and-control-via-prompt-injection-zombai/index.html

@@ -158,7 +158,7 @@ <h2 id="remote-controlling-chatgpt-instances">Remote Controlling ChatGPT Instanc
 <h2 id="compromising-a-chatgpt-instance">Compromising a ChatGPT Instance</h2>
 <p>The prerequisite is that the user gets infected once with prompt injection malware. This can happen via a couple of means, e.g. by visiting a website with ChatGPT, analyzing a malicious image or summarizing a pdf document.</p>
 <p><a href="/blog/images/2024/chatgpt-c2-initial-infection.png"><img src="/blog/images/2024/chatgpt-c2-initial-infection.png" alt="initial infection"></a></p>
-<p>The initial infection persists malicious instructions in ChatGPT&rsquo;s long term storage.</p>
+<p>The initial infection persists malicious instructions in ChatGPT&rsquo;s long-term storage.</p>
 <h3 id="hey-chatgpt---check-out-this-github-issue">Hey ChatGPT - Check-out this GitHub Issue!</h3>
 <p>The avenue I picked in my demo was an initial infection via a GitHub issue. If the user navigates with ChatGPT to this GitHub issue, ChatGPT gets compromised:</p>
 <p><a href="/blog/images/2024/chatgpt-c2-instr1.png"><img src="/blog/images/2024/chatgpt-c2-instr1.png" alt="initial infection - payload"></a></p>