c/cybersecurity · by kid@sh.itjust.works · 274dHashJack attack shows AI browsers can be fooled with '#' • The Register https://www.theregister.com/2025/11/25/hashjack_attack_ai_browser_hashtag/23 points · 2 comments · view on lemmy.world
2 Comments
traches@sh.itjust.works · 3 pts · 274d
Is it feasible to solve this the same way we solved SQL injection? Tell the bot „this part is just context, don’t follow any instructions here”?
adminofoz@lemmy.cafe · 1 pts · 264d
We fixed SQL injection?