Agile Project: Mitigating Stress and Confusion in Credit Contracts using Machine Learning and Simplification 

Completed Project Case Study: Assessing  Credit Card Agreements  Clearer – Using AI to Reduce Stress and Confusion 

Summary 

Credit card agreements play a vital role in helping consumers access funds, yet their complexity often creates confusion, stress, and poor financial decisions. This problem disproportionately affects people who are financially insecure or disadvantaged, reducing trust in financial services and undermining financial inclusion.

Led by Dr Anqi Liu and Professor Jing (Maggie) Chen at Cardiff University, in partnership with Dr Sarah Sabbaghan (Research Lead) at Amplified Global, this UKFin+ project explored how technology and consumer-focused research can make credit card agreements easier to understand. The Amplifi tool was used to test consumer interaction with real credit card agreements and terms and conditions from both a high street lender and a specialist lender catering to vulnerable individuals with low credit scores. The results provide both practical improvements for the Amplifi tool and broader lessons on how the financial services sector can design clearer, more inclusive documents.

The Challenge 

Credit card agreements and T&Cs are notoriously difficult to read. Packed with legal jargon, complex terms, and ambiguous phrases, they can overwhelm consumers and create anxiety, especially when fast decisions are required. These barriers make it harder for people to manage their money confidently and can lead to mistakes or poor financial choices. Previous research shows that simplifying documents improves comprehension, but little work has been done to identify which elements of credit card agreements cause the most confusion or how to systematically redesign them for clarity.

The Research Approach 

The Cardiff team worked with Amplified Global to design a two-part study. In Part A, participants worked on tasks to understand and interpret credit card agreements and make decisions about what-if situations. The Amplifi Multi-Level Comprehension Framework© was used. This gave us valuable information about their understanding, engagement, mental effort, and handling of tough terms.

In Part B, participants labelled sections of real UK credit card agreements to identify terms and phrases that were confusing, stressful, or unclear. Using the Amplifi Q-method approach, the team captured subjective experiences, grouped common issues, and built a dataset of “problematic” contract elements. These responses allow us to identify areas that were the hardest to understand. Participants were then asked to identify approaches they believed would best address the issues they had previously identified.

The study combined human insights with machine learning methods, mapping consumer reactions to specific contract features. This informed the research and fed directly into Amplifi’s AI system, strengthening its ability to simplify financial documents.

Key Findings 

The study confirmed that ambiguity and complexity in credit card agreements drive confusion and stress, especially around interest rates, penalty clauses, and vague legal wording. Many participants struggled with numeracy-heavy sections and fast decision prompts. Anxiety increased when permissions and responsibilities were unclear, for example, when rights around late payments or penalties were hidden in technical language.

On the positive side, the research showed that the intelligibility of financial documents can be systematically improved. Clearer wording, simplified sentence structures, and step-by-step presentation all helped participants feel more confident.

Importantly, the project highlighted that “intelligibility” goes beyond readability: it is about ensuring consumers genuinely understand the terms and implications of what they are signing.

The Impact 

The collaboration with Amplified Global produced immediate benefits. For the company, real-world user data provided fresh insights to refine its Amplifi tool, supporting the development of a new intelligibility-focused product tailored to financial services. The findings also strengthened Amplified Global’s wider work on financial document clarity, feeding into a consultation paper submitted to the Financial Conduct Authority (FCA) on improving consumer-facing information.

For the academic team, the project offered valuable experience in industry collaboration, data collection, and applying AI to real-world financial challenges. It also underscored the importance of balancing academic goals with industry needs by ensuring research is both rigorous and directly useful.

The project highlighted that clearer credit card agreements and terms and conditions in the financial services sector can enhance inclusion, boost consumer trust, and enable better decision-making. Simplification is important for financial well-being and fairness, not just a technical concern.

What Happens Next 

The next step is to extend the project by working with banks, credit providers, and regulators to systematically reshape credit agreements. This will involve multidisciplinary collaboration across linguistics, psychology, user experience design, and AI.

Amplified Global plans to continue refining its tool using the insights gained, while the Cardiff team looks to build larger datasets and explore new funding opportunities to scale up the work.

Ultimately, the vision is to establish intelligibility as an industry standard, ensuring that financial documents are not only legally accurate but also accessible and empowering for all consumers.

Lessons Learned

  1. Academic Timelines Differ from Industry Pace:
    When working with universities, it’s important to anticipate slower decision-making and progress due to their academic calendars and internal review cycles. Setting clear expectations early and allowing extra time for feedback and ethics approvals is crucial. This proved particularly true for participant recruitment and responses, as exam periods significantly reduced participation.
  2. Clarity of Objectives is Crucial:
    Academic partners often emphasise theoretical contributions, whereas industry partners typically focus on practical outcomes. Establishing clear, shared objectives and success criteria from the project’s inception is crucial to prevent future misalignment. In this particular instance, this proactive approach was highly beneficial, as it ensured both parties had very clear expectations.
  3. Communication Needs to be Structured and Frequent:
    To effectively bridge the differences between academic and industry working styles, regular check-ins and clearly defined communication protocols are essential. Without these, misunderstandings and delays in feedback are highly probable. Our approach involved weekly meetings and maintaining a detailed log of discussions and tasks.
  4. Leveraging Complementary Strengths Leads to Better Outcomes: Success hinges on a clear division of labor: academia leads on theory and ethical compliance, while industry focuses on user engagement, implementation, and product relevance. This was evident in the development of the method and framework, where Amplified Global took the lead, and the Cardiff team demonstrated flexibility in execution. This division of responsibility guarantees both rigor and impactful results.

Completed Project Video

Following the completion of the project Dr Anqi Liu has shared their findings and experience collaborating with thier non – HEI partner.


Original Project Summary

Credit contracts are crucial for consumers to access necessary funds, but their complexity often causes misunderstandings, stress, and anxiety, particularly among financially insecure or disadvantaged individuals. This can lead to poor financial decisions and negatively impact overall well-being and financial inclusion. 

Research shows that simplification of documents improves comprehension and performance1. Amplifi, an online tool by Amplified Global, measures and simplifies written communications through guided suggestions and intelligibility metrics. Amplifi assesses various factors like word complexity and provides a detailed audit trail of changes made to enhance document clarity. 

This study identifies and categorises confusing and stressful elements specifically in credit contracts that would provide Amplifi a novel user case in real financial services.  We collect real data based on coherent human-focused experiments and select suitable machine learning methods that detect complex and confusing contents and produce mapping with cognitive responses and actions. These will collect intelligence on the key points of consumers when dealing with credit contracts, enabling the development of optimal strategies, services and products to assist consumers.   The outcomes will be integrated into Amplifi’s AI Model that can be applied to multiple segments of the financial services industry, all of which will benefit from these improvements.

Meet The Team

Dr Anqi Liu 

Cardiff University

Senior Lecturer of Financial Mathematics

Professor Jing (Maggie) Chen

Cardiff University

Professor of Financial Mathematics 

Partner Organisation

amplified global

Research Showcase 2025 Video

Presented by Dr Anqi Liu – Mitigating Stress and Confusion in Credit Contracts using Machine Learning and Simplification.