Skip to playerSkip to main content
  • 4 minutes ago
Traditional credit checks can leave many women behind. Women in Mexico are often charged higher interest rates despite stronger repayment records. Fintech firms are using AI to challenge long-standing biases in lending.

Category

🗞
News
Transcript
00:00Luz Maria was trying to rebuild her finances at age 60, following a difficult few years.
00:06A family health crisis and the pandemic had strained her household income.
00:11She cares for her son, her mother and a sister with a disability.
00:15Luz Maria eventually settled her credit card debts.
00:18But banks continued to reject her applications for a new card, pointing to her credit history.
00:24First of all, I'm single and believe it or not, there's this idea here that if you don't have a
00:30man by your side,
00:31you're not responsible or you're not trustworthy.
00:36Her experience reflects a wider problem.
00:38In 2024, only 73% of women in Mexico had at least one formal financial product,
00:44like a savings account, a loan or insurance package.
00:48By comparison, about 81% of men in Mexico had one.
00:52At the same time, women were slightly less likely than men to fall behind on loan repayments.
00:57Yet they were still charged higher interest rates on average.
01:00So, what happens when lenders change the way they assess risk?
01:04Fintech companies like Story that offers credit products to underserved customers are increasingly using AI.
01:10But that comes with its own challenges.
01:14There is a risk that artificial intelligence could widen gender gaps by reinforcing stereotypes
01:20that do not benefit women.
01:26Part of that problem lies in the data used to train AI models.
01:30AI and fintech expert Daniela Camberos explains.
01:35When the fintech companies I worked for were using these models, it was clear that women were being sidelined.
01:42When they were given more opportunities, you realized that women were actually much better at repaying loans than men.
01:50Traditional credit systems focus on what a person owns or their credit history.
01:54But many women have less access to property or other assets that can be used as guarantees.
02:00Story says its models can instead look at additional information such as bills paid on time or spending patterns.
02:08When we incorporate a gender perspective, you consciously have to tell the artificial intelligence.
02:14When it's a woman, give extra weight to the alternative income she's going to mention and not to the credit
02:20bureau.
02:22This is where people play a key role in deciding what information the model uses, how it is weighted and
02:29whether the results need to be adjusted.
02:31For women like Luz Maria, this is an opportunity to be assessed on fairer terms.
02:37There are now some impressive investigative tools available so they can find out about the person, their track record, their
02:45payments and so on.
02:47And based on that, establish criteria that would allow them to be a little more confident in dealing with women.
02:54There are often times when women have several permeated and other passwords available.
02:56We'll see why some women, especially women can't end up in their own relationships.
02:57The first is who has been identified with women in travel,
02:59This is where women can share their own relationships with women.
02:59We know, there are a lot of different types of people in their own relationships.
03:00We are often different from the dude's friend.
03:00You

Recommended