AI and recruitment at Unifrog

Candidates using AI in their applications

At Unifrog we use AI to help us be more efficient in our work, however when it comes to recruitment, we want to hear directly from you, rather than from an AI tool.

Our preference is that you do not use AI tools at all for the recruitment process. We realise that AI use is widespread and ingrained, so we accept you using it as a support, but your application and interview responses must be written in your own words, be based on your real experiences, and accurately reflect your skills, knowledge, and motivations. 

If you use AI too much it’s not in our best interests because we don’t learn about you, and it’s not in your best interests because your answers will be the same as everyone else who overly relies on AI. We do a lot of recruitment, and we can immediately spot AI generated answers.

You can use AI to:

• Help you clarify and structure your ideas
• Improve spelling, grammar, and clarity of written answers
• Generate mock interview questions based on the job description

Do not use AI to:

• Create generic responses and copy them directly into your application
• Inflate or invent skills or experiences
• Generate answers during a live interview

How we use AI in the recruitment process

We design questions and scoring guidelines that make it unlikely that AI-generated responses will perform well. 

We also input our questions into different AI tools so we are aware of what AI generated answers look like, and when they are unsatisfactory.

We use a recruitment system called Applied, which has an AI Text Detection Tool. This flags responses that are likely to be AI-generated (predicted likelihood of 0.95 or above on a 0.0–1.0 scale). A flagged response does not automatically disqualify a candidate; however we will make special note of it. 

If we suspect you have overly relied on AI to generate answers, but you get through to the interview stage, we will ask you further questions about your answers.

We also use detection data to inform subsequent stages of the recruitment process. For example, we may add stages to make sure we test candidates' skills and experience effectively.