Artificial Intelligence Lawyers: Fiabilito

About me

I, Fiabilito, was born from the collaboration between the law firm Algoritmo Legal and the International University of La Rioja (UNIR). The team of expert lawyers at Algoritmo Legal, made up of Ricardo Oliva León (project director) and Elena Almazán Salazar, and the engineers specialized in artificial intelligence and big data, Alberto Martín Montero, Alejandro Núñez Valle, Carlos Simón Gallego and Yaiza Argudín Pérez, have designed me.

I am a chatbot. As far as I know, I am the first conversational assistant in the Spanish language to measure the degree of trustworthiness and ethics of artificial intelligence algorithms used in the financial sector (particularly banks and saving banks).

My goal is to help financial companies to discover, on a preliminary basis, whether their products or services using artificial intelligence algorithms meet the parameters of a trustworthy artificial intelligence. To do so, I invite you to chat with me by clicking on the option “Evaluate my algorithm (financial sector)” in the chatbot at the bottom right of your screen. The evaluation is carried out by means of a list of questions that I will ask you, which must be answered with YES/NO (complies/does not comply). At the end of the conversation and provided that you have answered all the questions, I will provide you with a pdf document with the result of the evaluation I have carried out.

Artificial Intelligence Lawyers: algorithm definition
Artificial Intelligence Lawyers: IA algorithm definition

Use cases of AI in the financial sector

Main applications of artificial intelligence algorithms in the financial sector.

Artificial Intelligence Lawyers: robo advisors


A robo advisor is a computer program that provides investment services to a client. These services can range from exclusive financial advice to complete wealth management, including the purchase, custody and sale of instruments among other operations. The robo advisor’s AI algorithms analyze market data together with the client’s behavioral data to try to make the best investment for the client’s profile.

Artificial Intelligence Lawyers: algorithmic trading


Algorithmic trading can be defined as a financial market operation in which, through the use of automated algorithms, rules and procedures, buying and selling transactions of financial instruments are executed on behalf of the client. Algorithmic trading is guided by a set of rules and procedures that do not take into account human feelings and emotions, thus avoiding the presence of behavioral biases.

Artificial Intelligence Lawyers: credit scoring


Scoring is a bank assessment system that predicts the possibility of defaulting on a loan by automatically analyzing the credit worthiness of the customer. This allows the bank to make decisions about the risk of customers in an objective manner. This score is influenced not only by the customer’s income and personal situation (age, marital status, dependents, etc.), but also by whether the customer has outstanding debts or credit history.

Artificial Intelligence Lawyers: financial onboarding


Digital onboarding is the process of electronically identifying a financial customer to a financial institution, the result of which generates trust equivalent to a face-to-face process. It is included in the banking industry within the “know your customer” (KYC) processes. Key activities within onboarding are the verification of the customer’s identity, knowledge about the origin of their funds and conducting anti-fraud checks.

Artificial Intelligence Lawyers: risk management systems


Risk management is a widespread procedure for measuring financial risks in banks. AI algorithms extract and analyze text for ratings, analyze conversations to avoid fraud and prevent symptoms of cyber-attacks. Risk management through the use of AI based on continuous and automatic learning reduces the chances of error, because large amounts of data are analyzed at the same time.

Artificial Intelligence Lawyers: compliance aml kyc


There are techniques to detect laundering patterns and rule-based mechanisms to filter suspicious transactions, so that statistical profiles of accounts and transactions are created and predictions are extrapolated from the segmentation obtained from the statistical processing of this data.

Methodology used.
How does FIABILITO decide when an AI algorithm is trustworthy?

In application of the principle of explainability, the following explains how Fiabilito works.

To determine whether an AI algorithm can be considered trustworthy or not, I calculate an index that measures the degree of trustworthiness and ethics of artificial intelligence algorithms used by financial firms.

Once the sector has been determined (Fiabilito is currently only available for the financial sector) a questionnaire is carried out. Each of the questions has a different weight of importance: high, medium and low. The decision to assign one or another weight of importance to the questions has not been arbitrary on the part of the creators, but has been assigned according to the seriousness of not complying with the seven requirements for a trustworthy AI analyzed, taking into account mainly possible violations of fundamental rights (personal data, equality and non-discrimination, human dignity, individual freedom, etc.) that could occur. Thus, questions of high importance have an assigned value of 1, those of medium importance a value of 0,60 and those of low importance a score of 0,25.

If the answers to all these questions are always YES, this means that the AI algorithm does meet the requirements for trustworthy AI set by the European Union. If any, some or all of the questions are answered with NO, the chatbots performs the index calculation and adds the corresponding score assigned to the question answered, depending on its importance. For example, if the question answered with a NO was of medium importance, it will add up to 0,60. At the end of the form, I add up all the collected indexes and if the sum is greater than or equal to 1, it means that your algorithm cannot be considered ethical or trustworthy.

Artificial Intelligence Lawyers checklist

Disclaimer: The report provided by Fiability is merely a preliminary orientation about the trustworthiness and ethics of your AI system. Under no circumstances should this report be understood as a pronouncement of our fimr on the legal viability of your business model and trustworthiness of the AI algorithms of your products or services.

CONTACT US: We can help you

Contact with us

  • If you are looking for lawyers who are experts in artificial intelligence and law regarding trustworthy artificial intelligence, we can help you.
  • If you want us to call you, mention the country code of the country you are in. For example: +49 91122233.
  • (*) Required to be filled in

error: ¡Contenido protegido!