Your Face as Your New Resume? Algorithm Will Assess Your Conscientiousness Based on a Photo

Hi!

Imagine this scene: you upload a new photo to LinkedIn. You think – “it turned out pretty good.” But before you even manage to gather the first likes, somewhere on a server, an artificial intelligence algorithm is scanning the features of your face. In a few seconds, it assesses your conscientiousness, openness to new experiences, and even your level of neuroticism. Based on this, the system recommends you (or not) to a recruiter looking for someone for your dream position.

Sounds like a scenario from another episode of “Black Mirror,” right? A team of researchers from the absolute top (Wharton, Yale, Indiana, Reichman University – Guenzel, Kogan, Niessner, and Shue, these names are worth remembering) has just published a study that could turn our thinking about recruitment and human capital upside down.

What did they discover? AI reads personality from faces – and does it surprisingly well

They took data from nearly 100,000 graduates of prestigious MBA programs. Using AI, they analyzed their profile photos from LinkedIn to “extract” personality traits according to the Big Five model (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism). They named this new indicator “Photo Big 5.”

And now for the best part: it turned out that this “personality read from a photo” predicts real-world outcomes in the labor market – from university rankings, through earnings, position held, industry choice, all the way to career path and promotions.

What’s more, the accuracy of these predictions is comparable to traditional metrics such as GMAT scores (the entrance exam for MBA studies) or the average grades from studies! But be aware – “Photo Big 5” only slightly correlates with these cognitive test results. This suggests that AI is capturing something completely different – a new dimension of potential that has been difficult to measure until now.

Huge potential, even greater temptation

What does this mean in practice? Instead of long, boring, and often unreliable (because who among us hasn’t tried to “answer correctly”?) personality questionnaires, a photo would suffice. Quickly, cheaply, on a massive scale. Access to an “objective” assessment of a candidate with a single click.

For companies that want to optimize processes and look for the “perfect fit,” this is like finding the Holy Grail of recruitment. For labor market researchers – a goldmine of data. Tempting, isn’t it?

Well, yes, but… the dark side of the force

This is where the serious difficulties begin, and questions that should keep us awake at night arise:

Discrimination on steroids? Remember the experiment with Heidi and Howard that I wrote about? Exactly the same skills, but different gender = different assessment. Even now, unconscious biases influence how we perceive people. What if AI, learning from our data (and our biases!), starts to systematically favor or discriminate against certain groups based on facial features? Will it be easier to “fix” bias in an algorithm or in the human mind? That’s a million-dollar question.

The end of privacy? Do we really want our face to become an open book from which every recruiter or potential boss can read our personality traits without our knowledge or consent? Where is the line between data analysis and hacking into our psyche?

Autonomy of the individual? Will we be judged based on something we have no control over – our facial features? Is that fair? Doesn’t this lead to a world where we are pigeonholed from the start?

My two cents – fascination and fear

I look at this research with a mixture of admiration for the possibilities of technology and genuine concern. On the one hand, the potential for a deeper understanding of what truly matters in a career is fascinating. Maybe this is a step towards a more meritocratic world?

On the other hand, the risk of creating a tool that will deepen inequalities and take away our control over how we are perceived is enormous. It reminds me of the debate around startups like Mercor, which automate recruitment using AI. Technology can be a powerful ally in eliminating human errors and biases, but it can also become a source of new, systemic problems.

What’s next? We need a serious conversation

This research is not some distant, academic curiosity. In my opinion, it’s a wake-up call. Before technologies like “Photo Big 5” become commonplace (read: in tools used by employers), we as a society, as an industry, as employees and employers, must have a fundamental discussion about boundaries, ethics, and regulations.

What do you think about this? Would you let an algorithm assess your professional potential based on a photo from LinkedIn? Do the benefits outweigh the risks? Where should the line be drawn?

I am very curious about your opinion. Let me know in the comments!