Mishcon’s AI Interviewer: The Future of Lawyer Recruitment?

Mishcon de Reya is deploying an AI chatbot system that interviews wannabe lawyers before they meet a human at the firm. Is this the future of lawyer recruitment?

How does it work?

Bright Apply is an AI-powered candidate screening solution developed by graduate careers platform Bright Network. It uses information from candidates’ applications to ‘start a conversation’. 

Instead of writing a long application form, candidates take part in a virtual interview where they can ‘expand on their experiences [and] motivations’, and why they want to work at this particular law firm. 

This process produces a transcript, with each transcript ‘carefully reviewed by the firm’s team’.

For now the firm stresses that its ‘Early Careers team continues to make all screening decisions’ and adds ‘the AI just helps us learn more early on; real people assess all the applications and make the decisions. While for the purposes of the trial we will be seeing how the AI scoring compares, we will not be making any decisions based on the AI scores’.

In short, for now this is a trial, and importantly humans are making the final decision.   

Is This A Good Thing?

Mishcon did a survey of some applicants and they say that ‘over nine out of 10 (93%) of student responses said the AI chat responses felt personalised and relevant to them’.

On social media the view was more varied. One Reddit poster wrote: ‘That was hideous and I was so excited about their firm. It took over three hours …. [and the AI] repeated so much of the same questions over and over. Wish they opened with the video part, at least so you didn’t feel so brain dead.’

Of course, in any application process you will get a range of feedback, that is inevitable. What is also inevitable is that law firms – which receive huge volumes of applications – will use tech to filter those applicants.

That said, predictable use of tech is not always a good thing. Should prospective lawyers be protected from such developments? Isn’t it vital for a human to meet a potential hire for such a complex job? Well, yes, and Mishcon stresses multiple times that this is just an initial screening and even at this stage humans are making the decisions. What’s not happening is the Mishcon team having to sit through hours and hours of face to face interviews at this early stage.

But, even so, law students may just say: ‘But why…?! Why change the way you do things?’

The firm explained: ‘Previously, we used application forms and one-way recorded video interview as a means to help us assess that large number of candidates….We think the chat-based interview is a fair, engaging and effective way of getting to know candidates.’

Also, they noted that aside from the above – and saving their team a ton of time – that this method ‘offers a more transparent and traceable process’.

But, then one might ask about bias. Is the system biased against certain types of applicant? Or perhaps certain answers unfairly get you rejected?

The firm said: ‘Bright Apply will statistically test the screening algorithms to ensure there isn’t bias in the system and use this analysis to refine and improve the AI scoring frameworks, guided by qualified and experienced occupational psychologists. Between them they have 40+ years’ experience in graduate recruitment, assessment, behavioural assessment, psychometric testing and recruiting technology.’

Of course, that is not a complete answer. And, as AL has explored many times, you cannot remove absolutely all bias from any system. The only thing you can do is be aware of it and perhaps agree that this bias is fair, given your business. For example, a person who continually uses profanities and slang in the AI interview will probably get rejected – AL would guess – but the firm would perhaps say such a bias was justified for a role as a lawyer.

That was an easy example, more subtle biases are much harder to both detect and justify. E.g. is there a bias perhaps against people who are more introverted? For example, many organisations may ask questions about teamwork, but some people who tend to be more introverted cannot honestly say that they would love the idea of working all day and night in a tight team, for weeks on end. Would that count against them? Or, would they just ‘know the script’ and provide ‘the expected answer’ to tell the AI system what it wants to hear? There don’t seem to be any easy answers to such questions.

Conclusion

AI interviews are perhaps inevitable, at least at the early stage of job applications, whether in the law or anywhere else. Are there risks that some candidates who don’t do well on such automated tests will be ignored? Yes. But, every type of interview method will tend to favour those who are cut out for them, and vice versa.

In short, all job screening systems are a compromise and a trade-off. None of them are perfect. This is perhaps simply ‘keeping up with the state of the art’ and an opportunity to reduce resources involved in the earlier stages of an application. Mishcon would likely add that it actually is ‘fairer’ as it gives applicants a good opportunity to provide expansive answers to many questions without having to sit through an initial face to face interview.

It seems likely that eventually – perhaps taking several years – that all major law firms will be using an AI system like this for recruitment.

–

Last word goes to Tom Wicksteed, Early Careers Manager at Mishcon de Reya, who said:  ‘We are really excited to be partnering with Bright Apply on this new tool, which will help us to get to know our candidates better at an earlier stage.’

‘The Bright Apply tool puts candidates at the heart of the process, giving them a better chance to show us who they are and what they are able to bring to our firm, at an earlier stage of the process. Our candidates will be able to tell us more about themselves, instead of being limited to a certain number of characters on an application form.

‘Not only does this give candidates an opportunity to detail their achievements, experiences and motivations but also helps our recruiters to make more informed decisions at an early stage of the process.’ 

—

You can see more about the system here.