How to write a resume with AI: tips and a ready-made prompt | Kaspersky official blog

The content explains that AI language models used for resume editing can inadvertently memorize and leak personal data, either due to training data retention, bugs (e.g., a Meta AI bug exposing conversations), or public indexing of shared chats (e.g., Grok). Users are advised to remove sensitive information before submitting resumes to AI chatbots and to disable AI training on their inputs when possible. Running AI models locally is recommended for handling sensitive data. The article warns that malicious actors may use leaked personal data to craft targeted phishing attacks and fake job offers. It also describes how ATS software, sometimes enhanced with AI, is used by recruiters to screen resumes, emphasizing that resumes must be optimized for both AI and human readers. Practical resume-writing tips are provided to improve chances of passing automated and human screening.

The primary impact is the risk of personal data exposure through AI chatbots and language models, which can lead to privacy breaches and enable targeted phishing and social engineering attacks. Users submitting resumes with sensitive information to AI services may inadvertently expose data to unauthorized parties or malicious actors. This can result in scams such as fake interview invitations, phishing links, or malware delivery. The use of AI in recruitment processes also means that poorly optimized resumes may be automatically rejected, affecting job seekers' chances.

Users should remove all sensitive personal information (e.g., name, contact details, company names) before submitting resumes to AI chatbots. They should review and adjust privacy settings on AI platforms, including disabling training on their inputs if possible. Prefer AI tools with clear privacy policies, HTTPS protection, and trustworthy ownership. Running AI language models locally is the safest option for sensitive data. Avoid using AI resume services from unknown developers. When writing resumes, tailor content for both AI screening and human readers to avoid automatic rejection. No official patch or fix applies as this is a privacy and usage guidance issue rather than a software vulnerability.