top of page

Beyond the Uncanny Valley: The Ethics of AI Disclosure in Recruitment



Recent advancements in AI have led to more natural and human-like interactions, potentially blurring the line between AI and human communication.


The uncanny valley, originally conceived for humanoid robots, describes the phenomenon where almost-but-not-quite-human representations can provoke unease or revulsion. In the context of conversational AI, we have reached a point where interactions are sufficiently natural that they no longer trigger this uncanny valley effect for many users. The natural implication of this is that users are now feeling more comfortable talking with AI systems, as we’ve already seen at companies like character.ai where humans have conversations with AI on average for 2 hours continuously.


Given this progress, it’s important to highlight the ethics of disclosure as it relates to recruitment, since in some interactions humans will not be able to know if they are talking to an AI system. Specifically, we believe that when engaging with a human, the following best practices should apply:


  1. Disclose that they are an AI or Virtual Assistant (“non-human”) at the earliest appropriate stage of the conversation.

  2. Explain the purpose and details of the conversation (similar to informed consent)

  3. Enable individuals to opt out of the conversation at any point

  4. Use visual cues wherever feasible to signify AI involvement



In terms of regulations, Colorado is leading the way and has stated in it’s recently passed Senate Bill 205, that after 2026, anyone deploying or developing AI systems that engage with consumers must disclose to each consumer who interacts with AI system that they are informed they are talking to an AI system, except in the case where it would be reasonable to assume that they would know they are talking to an AI system.


We expect to see similar legislation appear in other jurisdictions as AI systems have now crossed the uncanny valley.


At Kwal, we have these safeguards built into our AI system and whenever new legislation appears we take action immediately to update our systems to meet the requirements for disclosure and to ensure that candidates have a transparent experience.

コメント


コメント機能がオフになっています。
bottom of page