AI Cover Letters Are Changing How Employers Hire
Cover letters used to matter a great deal in job applications. A good letter connected a person's resume directly to the job requirements. It showed employers which candidates had real potential, especially those with unusual career paths. Most importantly, it signaled that the applicant cared enough to put in serious effort. They weren't just sending out mass emails with a salary demand.
The Rise of AI-Generated Applications
That traditional system is breaking down. Large language models have completely changed the game. Job seekers can now create perfectly tailored cover letters instantly. With a single click, anyone can present themselves as a meticulous and dedicated applicant. They can do this hundreds of times every single day.
A new academic paper by Anaïs Galdin from Dartmouth College and Jesse Silbert from Princeton University examines this shift. The researchers used data from Freelancer.com, a popular jobs platform. They compared activity from before and after ChatGPT became widely available. Their findings reveal two major trends.
Cover Letters Have Gotten Much Longer
First, cover letters have significantly increased in length. Before the era of large language models, the median cover letter on Freelancer.com was 79 words long. It's worth noting that this platform focuses on one-off tasks, so letters were typically more concise than those for full-time positions.
A few years after ChatGPT's release, the median length rose to 104 words. Then, in 2023, Freelancer.com introduced its own built-in AI tool. This allowed users to craft proposals without leaving the website. The applications written using this tool—which can be definitively identified as AI-generated—are even longer. Their median length is 159 words. That's more than double the original human-written baseline.
Employers Have Stopped Paying Attention
The second major finding is that companies no longer value the content of these letters. When only some applications showed genuine effort, those letters were worth reading carefully. They likely came from the best workers. Now, when every letter appears polished and detailed, there is little point in reading any of them closely.
To demonstrate this, the researchers used AI to score every cover letter. They looked at nine different categories. These included evidence that the applicant actually read the job advertisement and the ability to write clear English. Each category received a score from zero to two, with a maximum total of 18 points.
Before AI tools became common, the median score was 3.9. After their adoption, the median score nearly doubled. This uniformity has real economic consequences.
Wages Are Falling and Hiring Quality Is Dropping
In the pre-AI era, a well-written proposal on Freelancer.com was worth an extra $26 per task. This was a significant amount on a platform where the median task paid around $100. After AI cover letters became ubiquitous, this premium completely disappeared.
Galdin and Silbert estimate that wages on the platform are now 5% lower than they would be without AI-generated letters. Hiring rates are also 1.5% lower. Employers have lost a reliable method to distinguish strong candidates from weak ones. As a result, they have reduced pay for all new hires. They also more frequently end up recruiting less suitable candidates.
For business owners, the decline in candidate quality is offset by the reduction in wages. However, the researchers calculated that the overall benefit to companies is smaller than the total losses suffered by workers. The labor market is becoming less efficient for everyone involved.
