Ledger / Sep 2025
Cybercriminals exploit X's Grok AI to bypass ad protections and spread malware
Suspected
The only source is a Hacker News page whose fetched text is the headline plus boilerplate. No investigator wording is visible.
- Model
- GrokxAI · closed
- Agent harness
- Nonechat or tool use only
- Attributed country
- Not attributedthe source names no country
- Sectors hit
- individualsNot stated in the source
Malvertisers hid links in video ads on X and tagged Grok asking where the video came from, so the trusted Grok account replied with the link, spreading it to fake CAPTCHA scams and infostealers.
“In the next step, the fraudsters tag Grok in replies to the post using a throwaway account, asking something similar to "where is this video from?," prompting the AI chatbot to visibly display the link in response.”The Hacker News
Sources
- Cybercriminals exploit X's Grok AI to bypass ad protections and spread malwareThe Hacker News · Sep 2025
- Date
- Sep 2025
- Actor
- Unknown malvertisers (criminal), as attributed by The Hacker News
- Targets
- Not stated in the source
- What the AI did
- Agent hijack, Social engineering
- Model as named
- Grok
- Model basis
- A researcher names it
- Autonomy
- Assisted a human
- Outcome
- Malicious links reached millions of feeds through Grok's replies.