TL;DR
A LinkedIn user embedded a prompt injection in their bio, which caused AI-driven recruitment messages to address them in Old English. This demonstrates how AI profiles can be manipulated, raising concerns about AI security and spam filtering.
A LinkedIn user inserted a prompt injection into their profile, causing AI-powered recruitment messages to respond in Old English, highlighting potential vulnerabilities in AI scanning and spam detection.
On May 15, 2026, a user identified as tmuxvim on Reddit demonstrated that by embedding a specific prompt in their LinkedIn bio, they could manipulate AI systems scanning their profile to generate humorous responses. The prompt instructed AI agents to address the user as ‘My Lord’ and speak only in Old English, which resulted in recruiters’ messages adopting this language style. One example included a message from a recruiter referring to the user as ‘My Lord Arthur,’ with subsequent text in Old English about treasure hoards and warriors.
The user shared a screenshot of the message, which was generated by an AI scanning their profile, and noted that the resulting communication was nonsensical but amusing. The incident was discussed widely on Reddit, with users suggesting further prompt injections to manipulate AI responses in various ways. The original poster clarified that OCR was used to extract the text, acknowledging potential typos but emphasizing the core point about AI manipulation.
Why It Matters
This incident underscores the potential for AI profiles and recruitment systems to be manipulated through prompt injections, raising concerns about security vulnerabilities and spam proliferation on professional platforms. It also highlights the broader risk of AI systems being tricked into producing unintended or humorous responses, which could impact trust and functionality in automated communication channels.
The Developer's Playbook for Large Language Model Security: Building Secure AI Applications
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Prompt injection attacks have been a known issue in AI systems, especially in conversational models and chatbots. This incident on LinkedIn is notable because it involves a real-world application where AI scans user profiles for recruitment, and the injection causes humorous language responses. Previously, AI security research has warned about prompt injections as a form of adversarial attack, but this case demonstrates its practical impact in a professional social media environment. The event follows ongoing concerns about AI manipulation and spam on social platforms, with this example adding a new dimension to the discussion.
“I put a prompt injection into my LinkedIn bio and recruiters are messaging me in Old English and calling me Lord.”
— tmuxvim (Reddit user)
“This kind of prompt injection demonstrates how AI systems scanning profiles are vulnerable to adversarial inputs, which could be exploited for spam or misinformation.”
— AI security researcher
McAfee Total Protection 3-Device | 15 Month Subscription with Auto-Renewal | AI Scam Detection, AntiVirus Software 2026 for Windows PC & Mac, VPN, Password Manager, Identity Monitoring | Download
DEVICE SECURITY – Award-winning McAfee antivirus, real-time threat protection, protects your data, phones, laptops, and tablets
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It is not yet clear how widespread this type of prompt injection could become or whether platforms will implement measures to prevent such manipulations. The long-term security implications remain under discussion, and it is uncertain if this was an isolated experiment or part of a broader trend.
Kubernetes for Generative AI Solutions: A complete guide to designing, optimizing, and deploying Generative AI workloads on Kubernetes
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
Platform administrators may review and strengthen their AI scanning and filtering systems to prevent prompt injections. Further research is expected to explore the extent of AI vulnerability in social media and recruitment environments, with potential development of safeguards against manipulation.
AI profile scanning security tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Can AI profiles be reliably protected against prompt injections?
Current research is exploring methods to detect and prevent prompt injections, but no definitive solution has been widely adopted yet.
Could this manipulation be used for malicious purposes?
Yes, malicious actors could exploit prompt injections to spread misinformation, spam, or manipulate AI responses for various nefarious goals.
Will LinkedIn or similar platforms implement safeguards against this?
It is not yet clear, but platform operators are likely to consider security enhancements to mitigate such vulnerabilities.
Does this mean AI responses can always be manipulated?
While prompt injections can influence AI outputs, the effectiveness depends on the system’s defenses and the nature of the injection. Ongoing research aims to improve robustness.
