Harnessing AI: Transforming Recruitment with Generative AI and Cutting-edge Tools

Generative AI has moved from novelty to daily habit for recruiters in barely a year, and the shift shows up in how fast the tools themselves keep changing. A live poll of the audience checks where people actually stand: who's using ChatGPT, who has hooked GPT into an ATS or CRM, who drafts emails from AI-generated text, and which platform — ChatGPT, Claude, or others — has become their default.

From there, the conversation turns to what the newest tools can do. Image generators like Ideogram handle text inside graphics better than older options. A tool called Suno writes full custom songs from a text prompt in under a minute. Video and avatar tools have advanced to the point where a cloned voice and face can deliver a scripted message that looks live, raising the question of how far off it is before a recruiter's own AI avatar could record custom video messages automatically.

That same technology cuts both ways on the candidate side. Applicants are already using AI to rewrite resumes to match job postings, mass-apply across postings, and in some cases feed answers into assessments and interviews. Some companies have started asking candidates to sign statements confirming they are who they claim to be and won't use AI during interviews. States including New York and Maryland have passed or proposed laws requiring disclosure of AI use in hiring and consent for facial recognition, with more legislation moving through California and at the federal level.

Real-world experiments get walked through in detail, including:

• An early AI internship experiment where AI "interns" handled repeatable design and content tasks well but struggled with anything new
• A customer service AI deployment that replaced hundreds of human agents while handling millions of conversations faster than the team it replaced
• Autonomous coding agents capable of debugging, browsing, and taking freelance gigs on their own

Practical, hands-on building rounds things out: creating a custom GPT trained through conversation to produce a source-of-hire metrics dashboard complete with tables and staffing insights, and building an image-generation GPT to produce consistent illustrations of different engineering roles for a student outreach project — including correcting the tool when its early outputs lacked diversity.