Will generative AI kill the resume?
After its invention more than 500 years ago, the resume has mutated several times, the latest and not the least: the arrival of the internet and the digitization of this piece of paper so useful. But since 2023 and the introduction of generative AI to the general public with ChatGPT, the usefulness of this tool inseparable from the job search is beginning to be questioned. Anyone can now generate dozens of quality resumes and cover letters hyper-adapted to each targeted company with a single click. What to do to determine the real skills and real motivation of candidates?
Companies must reinvent their recruitment processes to go beyond the classic resume. What if the solution came from generative AI voice assistants, capable of having hyper-realistic conversations with candidates? A promising track, but which also raises many ethical and human challenges.
Ready to discover the future of the resume? Follow the guide!
The resume, an essential recruitment tool
History of the resume
Okay, let’s go through the history box: where does the resume actually come from? Well, let’s start with the origins of the term itself. “Resume” comes from the French word “résumé”, which means “summary”. The term “Curriculum Vitae”, or “CV” for short, is also sometimes used in the US, particularly in academic contexts. It comes from Latin and literally means “course of life”. This Latin expression was already used in the Middle Ages by universities to designate a student’s academic path, but not yet for the job application document we know today!
Here is a timeline from the first known resume to the present day:
- 1482–It was Leonardo da Vinci himself who wrote what is now considered the first resume. In a letter addressed to the Duke of Milan Ludovico Sforza, he detailed his (numerous) talents as an engineer and artist to get a job. An ancestor of the resume that lays the foundations: to present oneself to a potential employer to arouse his interest. And Leo did arouse his interest, as he spent 17 years in the service of the Duke! One resume sent, one hire, nice perf! Well, knowing the character, we’re not surprised.
- 1914 — In his book “How to succeed in business” released in 1914, an American named James Libbey began to formalize the basic rules of the modern resume. With the well-known and now standard sections: “experience”, “education”, “skills”. With the foundations laid, it was in the post-war period and the associated economic boom that the resume became essential for applying to large companies. Recruitment firms and resume databases developed, further accelerating its adoption and making it essential.
- 2000 — The Internet arrived in force and gave the slightly aging resume a facelift. It was the gradual end of the paper resume and physical resume databases. Job boards and LinkedIn finished the job and the online resume became the norm. Make way for databases and resume scoring. For better or worse, recruiting entered a new era, that of Big Data.
- 2023 — Boom! The transition from paper to online resume barely finished, another change is in store for the resume: In October 2023, ChatGPT was released, introducing generative AI to the general public. A major change awaits the resume: we can now generate ultra-personalized resumes and cover letters in a few clicks. Enough to question the very relevance of the resume as a recruitment tool…
So, after more than 500 years of good and loyal service, is the resume about to retire?
The resume, a key tool for recruiters: meeting, motivation, comparison
But what is a resume for? Its primary role is to allow a meeting and a representation between a candidate and a recruiter, without them needing to be available at the same time. Without the resume, a candidate would have to go door-to-door to companies, hoping to arrive at the right time to talk to the recruiter. Tedious and inefficient… With a resume, the candidate and recruiter no longer need to be in the same place at the same time. The candidate sends this precious document by mail, email or LinkedIn and the recruiter can study it when they have time. The resume allows to pass a first step, to arouse an interest, before obtaining a face-to-face interview.
But the resume also has another key function: it gives an overview of the candidate’s motivation. Taking the time to write a resume and cover letter targeted to the company shows one’s interest in the position. A valuable element for the recruiter.
The resume is also a great comparison tool. By standardizing the presentation of profiles, it allows the recruiter to quickly compare the experiences, education and skills of different candidates. Of course, this also has its limits, but it must be noted that the resume facilitates the selection work.
The resume in the era of generative AI
The resume has survived many revolutions: from the democratization of the personal computer to the advent of the Internet… But with the rise of generative AI and tools like ChatGPT, it’s a whole different story. Because for the first time, the very relevance of the resume is being questioned.
Of course, it has always been possible to embellish or even falsify a resume. Some even estimate that most job seekers have already lied on their resume. Recruiters have tools like background checks or trick questions in interviews to catch the big bad liars.
What really changes here is the magnitude of the phenomenon and the quality of the deception: With generative AI, each candidate is able to create dozens of resumes and cover letters in record time, targeting dozens of different companies. No more spending hours adapting your resume, the AI takes care of it for you. Just enter some basic information (experience, education, skills), select the position and company you’re targeting, and you’re done. In a few seconds, the AI generates an ultra-personalized resume and cover letter, full of keywords and phrases tailored to the job offer.
But then, how can recruiters make sense of it all? How to distinguish the really interesting and motivated profiles in this mass of artificially generated applications? Because the risk is great of drowning recruiters under resumes created on an assembly line, without reflecting the quality or real motivation of the candidates.
Let’s take a concrete example. Pauline, a young graduate in marketing, is looking for her first job. Using a generative AI tool, she creates dozens of targeted resumes in a few clicks for positions as a product manager, community manager, communication manager… Positions certainly related to her training, but for which she does not necessarily have a real appetite. Yet, on paper, her profile seems to match perfectly. She receives many interview proposals, but struggles to convince in interviews, because her resumes do not reflect her deep aspirations. A waste of time for her and for the recruiters.
Another problem: by relying on pre-established models, the AI will tend to highlight the skills “expected” for the position, and not the real skills of the candidate. Let’s take Pauline’s example again. To generate her product manager resumes, the AI will draw on the skills most often cited in job offers: project management, market analysis, product strategy definition… Skills that Pauline, a recent graduate, does not yet fully master. But no matter, the AI will highlight them, giving a distorted image of her profile. The resume then becomes a mere reflection of market expectations, far from the singularities of each candidate.
And this is just the beginning. With the progress of AI, we can imagine that intelligent agents will soon be able to apply completely autonomously on behalf of their “master”. No more need for human intervention, the AI will take care of finding job offers matching the profile, generating resumes and cover letters, and even responding to the first exchanges with the recruiter. A dizzying prospect, which raises many ethical questions.
The observation is clear: in the era of generative AI, relying solely on the resumes and cover letters received no longer makes sense for effective recruiting. Companies must reinvent their approach to distinguish the true from the false, the human from the machine. This is a major challenge to avoid missing out on real talent, drowned in an ocean of AI-generated applications. But then, what solutions?
We could have the idea of an AI to sort through the growing number of resumes generated themselves by AI? But AI on top of other AI to detect AI, it sounds a bit off in this context, doesn’t it?
I could tell you about the decentralized resume on the Blockchain, this public platform where everything is traced and validated by the organizations themselves. For example, a candidate could no longer lie about their experiences because they would all be digitally “stamped” and validated directly by their former employer, making this resume standardized and unfalsifiable.
But, although passionate about this Blockchain technology, I will not dwell on the subject. If you are interested, I invite you to download for free the Blockchain and HR report that I wrote with my partners at Tomorrow Theory. A reference and a world first on the subject of 250 pages, in collaboration with Mazars and the inputs of ADEME.
No, my reflection today focuses on a simpler application of generative AI to implement at the moment: voice assistants.
Generative AI-powered voice assistants to replace the resume?
Faced with the limitations of the resume in the era of generative AI, companies must innovate to reinvent their recruitment processes. What if the solution came from generative AI voice assistants, these new interfaces that could revolutionize the work of recruiters?
Voice assistant?
What is a voice assistant? You all know Siri or Alexa, the assistants of Apple and Amazon. Imagine these assistants but with a voice and an interaction capacity indistinguishable from a human. Here the image will speak better than words, watch the video above before continuing reading!
For HR professionals, these voice assistants represent a unique opportunity to rethink the candidate experience, while gaining in efficiency. Imagine: instead of spending hours combing through resumes that are often unrepresentative, recruiters could rely on authentic conversations, conducted by their voice assistant. A considerable saving of time and quality in the pre-selection of candidates.
Better yet! The voice assistant could in a way dynamically build the resume of each candidate internally, and thus standardize the internal candidate data.
Echoes of Tomorrow
Concretely, here’s how it could happen. Pauline, our young marketing graduate, applies to “Assurancia”, a large insurance company. Instead of submitting a classic resume, she is invited to chat with “Victor”, Assurancia’s voice assistant.
From the first minutes, Pauline is pleasantly surprised. Despite the somewhat “traditional” image of the insurance sector, Victor adopts a modern and dynamic tone.
“Hello Pauline, this is Victor, the AI representing Assurancia! I’m delighted to make your acquaintance. At Assurancia, we are convinced that innovation starts first and foremost with talent. Tell me a little about your background and what motivates you to join our company!”
Throughout the conversation, Victor asks targeted questions about Pauline’s experiences, her digital marketing skills, but also her interpersonal skills and team spirit. Pauline feels considered in her entirety, beyond the lines of her resume. Victor also gives her information about Assurancia’s values, its commitment to continuous training and internal mobility.
After 30 minutes of discussion, Pauline feels like she has had a real interview, without the pressure of face-to-face. She was able to show her motivation and personality, while having a clearer vision of what her role at Assurancia would be.
On the Assurancia recruiters’ side, the assessment is just as positive. By analyzing the report generated by Victor, they discover a candidate with promising potential, with a real sense of customer service and an appetite for innovation. Key qualities for the position of digital communication manager, which would not necessarily have appeared on Pauline’s resume.
Convinced, the recruiters decide to invite Pauline for a more in-depth face-to-face interview. They know they can get straight to the point, by delving into the points identified by Victor. This is a considerable saving of time and efficiency, without sacrificing the quality of recruitment.
The 4Cs
For candidates, it’s an opportunity to develop and showcase the famous 21st century skills, the “4Cs”: Creativity, Communication, Collaboration and Critical thinking. Key skills in a constantly evolving world of work, where adaptability and emotional intelligence are more sought after than ever.
Instead of sending dozens of resumes, candidates would go through dozens of pre-interviews with various companies’ AI voice assistants, allowing them to develop their 4Cs and prepare for the real interview, face-to-face with a human this time.
During her exchange with Victor, Pauline was able to demonstrate her creativity by proposing original ideas to animate the community on social networks. She showed strong communication skills by adapting her speech to her AI interlocutor. She showed her collaborative spirit by mentioning her associative experiences, where she was able to federate teams around common projects. Finally, she demonstrated critical thinking by finely analyzing the challenges of digital marketing for the insurance sector. And most of all: she proved she was human (for now!).
These are all skills that are difficult to assess on a classic resume or cover letter, but that make all the difference in a professional situation. With voice assistants, candidates have the opportunity to highlight them in a concrete and authentic way.
HR at the heart of this transformation
Of course, for this system to be effective, HR teams have a crucial role to play. They are the ones who will “train” the voice assistant, by defining the evaluation criteria, the questions to ask, the expected answers, the “physical” appearance of the avatar. A real “design” work of the candidate experience, where every detail counts. HR will also have to ensure that the voice assistant complies with regulations on non-discrimination and protection of personal data.
AI tools are not exempt from risks of bias and discrimination. Indeed, these software are trained on historical recruitment data that may themselves be biased. If in the past, certain profiles have been favored (men, graduates of top schools…), the AI risks reproducing and even amplifying these stereotypes.
To avoid this, HR will have to ensure the diversity and representativeness of the data used to train the AI models. A real challenge at the moment when NONE of the 10 most used AIs in the world have yet published their training data or how they are educated to respond to users. HR will also have to put in place control and audit mechanisms to detect possible biases. AI should remain only a decision support tool, with human oversight to validate recommendations.
Another challenge for recruiters: learning to collaborate with these new AI “colleagues”. It’s not about being replaced by the machine, it’s about replacing the writing and sending of resumes! A pre-interview acting as a filter. Voice assistants can handle repetitive and time-consuming tasks (pre-selection, interview scheduling…), allowing recruiters to focus on higher value-added activities (fine candidate evaluation, new talent integration…).
This is a paradigm shift for HR professionals, who will have to develop new skills. Understanding how voice assistants work, knowing how to train and supervise them, interpreting their recommendations… So many new skills to acquire to make the most of these tools. It sounds technical put like that, but don’t worry, all this will be made easy, the hardest part will be to get started.
This vision is not far from becoming a reality, companies have been using automated online video recordings in their recruitment process for years, but with the rapid progress of conversational AI, the pace is accelerating! Pioneering companies like Unilever or L’Oréal are already experimenting with voice assistants to pre-select candidates, with promising results in terms of diversity and candidate experience.
Ethical and legal considerations: The European AI Act
The use of AI, and in particular chatbots, in recruitment processes inevitably raises questions about the collection and use of candidates’ personal data. The future European regulation on AI, the AI Act, which has just been voted in the European Commission, aims precisely to frame these practices.
Four levels of risk defined by the AI Act
The AI Act thus distinguishes 4 categories in AI use cases:
- Unacceptable risk: prohibited AI practices because they infringe on fundamental rights, such as social scoring by governments or predictive justice.
- High risk: AI systems used in sensitive areas (education, employment, credit, justice…) and subject to strict obligations.
- Limited risk: AI systems with specific transparency obligations, such as chatbots.
- Minimal risk: AI systems authorized without particular constraint, such as anti-spam filters.
Implications for AI recruitment
In the case of recruitment, the use of AI chatbots to interact with candidates would a priori fall under the “limited risk” category. This implies transparency obligations: candidates must be informed that they are interacting with a conversational agent and not a human recruiter.
However, some uses could switch to the “high risk” category, for example if AI is used to make automated hiring decisions, without human supervision. The obligations would then be much more restrictive: risk assessment, technical documentation, registration in a European database, implementation of human surveillance measures…
Prohibited practices and ethical responsibility
In the field of employment, the AI Act prohibits certain practices considered too intrusive, such as analyzing candidates’ emotions (with very limited exceptions). Even if the temptation may be great to exploit all the possibilities offered by AI to “decode” candidates, companies will therefore have to stick to reasonable and proportionate uses, and that’s a good thing!
Beyond strict compliance with regulations, it is also their ethical responsibility. Candidates must remain in control of their data and image. This requires transparency about AI recruitment processes, the obligation for AI to present itself as such to the candidate, and the possibility to not submit to it and to request the deletion of one’s data.
Conclusion
So, is the end of the resume near? In its current form, most likely! This recruitment staple is shaken to its very foundations by the emergence of generative AI. But rather than seeing it as a threat, let’s seize this opportunity to deeply reinvent our recruitment processes.
With AI voice assistants as allies, we can give oral interaction a central place again in the recruitment process. Admittedly, the dialogue is with a machine, but it’s an opportunity for the candidate to highlight their personality and their famous 21st century skills, the “4Cs” that make the difference in a company. AI could become a facilitator, a revealer of human potential, before the real exchange with the recruiter.
Of course, this revolution raises many ethical questions. How to ensure that the human remains at the center of the recruitment process, despite the rise of human-machine interactions? How to frame the collection and use of personal data, to avoid the excesses of an AI that would excessively profile candidates?
Faced with these challenges, HR professionals have a crucial role to play. They are the ethical guardians of their organization, the ones who must define the limits in the use of these technologies. It is up to them to tackle these subjects head-on, to invent the recruitment of tomorrow today.
But AI is not alone in reinventing the resume. The blockchain, with its certified and anonymous resumes, or extended reality, with its immersive simulations, are all promising avenues. It is probably a convergence of these technologies that will shape the future of recruitment. It’s up to HR to act as conductors to harmonize these innovations in the service of the human. An exciting challenge, which will require combining strategic vision and operational agility. The future of the resume will be technological, yes, but more than ever, it will have to remain human!
If you want to know more about Gen AI and the future of work, check out the book I co-authored with Jeremy Lamri — Generative AI at work: Reshaping the future of society
[Article written on June 13, 2024, by Gaspard Tertrais with the support of the Claude 3 Opus algorithm for approximately 20%. Main image created with DALL.E-3, 2024].
