Credit scores can be inaccurate due to errors and biases
A woman's mortgage application was denied due to a poor credit score despite having an excellent credit history.

A significant number of people rely on their credit scores to secure loans and financial services. However, these scores are often calculated using complex algorithms that involve data from various sources, including credit reports and payment history. The accuracy of this information is not always guaranteed, as evidenced by the case of Sophie Hope.
Hope's experience is a stark reminder that credit scores do not necessarily reflect an individual's true financial situation. In her case, her mortgage application was denied due to a "very poor credit score, despite having an excellent credit history up until then. Upon investigation, it emerged that someone had opened accounts in her name on various online catalogue sites.
The damage to Hope's credit score was severe and long-lasting, with debt accumulating over three to four years before she even knew about it. The experience left her shocked and wondering how such a situation could occur without her knowledge or consent. Her story highlights the potential for errors and inaccuracies in credit scoring systems.
Hope's case raises important questions about the role of artificial intelligence (AI) in credit scoring, particularly when it comes to making critical decisions about an individual's financial worthiness. As more lenders rely on AI-driven systems to assess credit risk, concerns are growing about the fairness and accuracy of these algorithms.
Credit scores are often seen as a key factor in determining our ability to borrow money, but they can be flawed and biased towards certain groups of people.
The UK has three major firms - Experian, Equifax and TransUnion - that produce credit scores based on similar data sets. This raises questions about the accuracy and fairness of these scores.
Critics argue that traditional credit scoring systems are not forward-thinking, instead relying on past financial behavior to make predictions about future creditworthiness. They point out that this can disadvantage young people and recent immigrants who have limited or no credit history.
The use of artificial intelligence in lending decisions is becoming more widespread, with some companies offering alternative ways for lenders to assess a client's credit worthiness, potentially providing more nuanced and accurate assessments.
The increasing use of artificial intelligence in lending decisions raises questions about its fairness and accuracy. Some companies believe AI can make more loans available at lower prices to a wider range of borrowers by analyzing a large number of variables.
Mortgage expert Sarah Tucker agrees that AI has the potential to help lenders analyze complex information, identify patterns quickly, and build a more detailed picture of an individual's affordability and risk. If technology allows for better understanding of so-called thin-file applicants, rather than relying on rigid rules, there is a real opportunity to improve financial inclusion," she believes.
However, concerns surround the use of AI in lending decisions. Consumers need to understand what information is being used to make decisions about them and have the ability to challenge any errors, according to Tucker. This echoes the views of Sandra Wachter, a professor at the University of Oxford who has expressed similar reservations about AI's potential to replicate existing social inequalities.
Existing regulations and new guardrails are crucial for preventing AI systems from perpetuating historical biases in evaluating credit worthiness, according to experts. This is a necessary step towards true financial inclusion, rather than simply relying on technology as a tool for change.
Firms must establish clear guidelines and oversight mechanisms to ensure that AI-driven decisions do not fortify or enhance existing discriminatory practices. Without such safeguards, the risk of perpetuating biases in the system increases significantly. The goal should be to create more accurate and fair evaluations over time.
A company that has developed an automated credit evaluation model claims to have made significant strides in addressing bias through its model training process. This involves incorporating objective evaluations designed to identify and mitigate bias, allowing the system to evolve towards greater fairness.
Evangeline Atkinson's frustrating experience with her credit score led her to establish a new credit reference agency in 2021, alongside her childhood friend who had also struggled with getting credit.
The new credit reference agency Noggin HQ has been authorized by the FCA in June, marking a significant development in the industry. Atkinson's firm is designed to help consumers access affordable credit by analyzing their current account transactional data.
To achieve this goal, Noggin HQ helps lenders interpret thousands of individual data points about an individual's spending habits, providing a more nuanced understanding of their financial situation. This approach is notable because it avoids relying on AI, which Atkinson has argued can introduce inherent biases into the system.
The use of AI in credit scoring has become increasingly prevalent, with even the big three credit scoring firms experimenting with its potential applications. Experian and OpenAI have partnered to allow users to check their credit score within ChatGPT, while Equifax has launched a platform called Equifax Amplify AI that uses machine learning for sharper credit assessment and fraud detection.
TransUnion is also incorporating Google's Gemini models into its proprietary analytic data assets, as revealed in March. This shift towards AI-powered credit scoring may signal a broader move away from traditional methods, according to consumer finance lawyer Funmi Olufunwa, who notes that companies are now exploring alternative approaches to evaluating creditworthiness.
Traditional credit scoring methods are being reevaluated by companies as they search for more effective predictors of future financial performance.
This shift may be driven in part by younger consumers who are accustomed to sharing their data and see AI-powered credit scoring as a natural progression of the digital world they've grown up in, with its emphasis on technology-driven services.
Facts based on reporting originally published by BBC News Technology.
You may republish this story, in full or in part, if you credit Noti Group and link to it (licence CC BY 4.0). Photos are not included.












