What are potential ethical problems that must be addressed?
Consulting Offering
HSO's AI Briefing
Alongside HSO experts, uncover the potential and shape your AI adoption journey for success.
Get started
Understanding the Ethical Risks of AI in Financial Services
10 Aug, 2023
How does AI work?
Understand some of the potential ethical issues AI poses for financial services institutions and what technology leaders have agreed to provide as safeguards.
Artificial Intelligence (AI) has become one of the most talked about technologies these days, and for good reason: as the technology matures, it is finding application across the entire spectrum of human endeavors, from science to art…even financial services.
AI emulates human decision-making by consuming vast amounts of data and applying algorithms that arrive at an appropriate decision based on past good and bad decisions. The algorithm is tested by applying new data and measuring the machine’s responses.
Over time, the model “learns” the difference between a good decision from a bad decision based on the algorithm and live human interaction, offering feedback about the quality of its decisions. At the same time, AI designers are constantly adjusting the learning algorithms and models to ensure optimal responses.
AI is currently being used in a number of financial services applications, including managing asset allocations, assessing portfolio risk, qualifying loan applications, and automated chatbots that recommend services based on client needs. The benefits of AI promise to help financial services companies reduce costs, maximize client portfolio values, and accurately assess the creditworthiness of loan applicants.
There are significant benefits to employing AI within financial services such as:
- personalized customer experience,
- risk management and fraud detection,
- enhanced data analysis, and
- operational efficiency and cost reduction.
But with great power comes great responsibility.
- 1
Transparency
It is nearly impossible to understand how complex AI algorithms work, particularly proprietary algorithms from third-party developers. This lack of transparency can be problematic for regulators, clients, even the company itself, to understand why a questionable transaction was made.
- 2
Accountability
Along with the potential lack of transparency, complex decision-making algorithms can make it difficult to assign responsibility and hold an entity accountable if errors or mistakes occur.
- 3
Privacy
AI learns by assimilating vast amounts of data. This can involve inadvertently including sensitive or personal information in the financial services industry. Data privacy and security is critical to prevent misuse and unauthorized access to sensitive information.
- 4
Bias
Created by feeding data to a model, the algorithm is affected by the bias of the data going into it. This can take many forms, such as a predisposition to specific asset classes or trading strategies. Bias can also take the form of discriminatory practices applied to marginalized demographic groups when evaluating loan or insurance applications.
- 5
Misplaced Dependence
AI is a tool for decision-support. Placing too much reliance on the tool without sufficient human oversight can mean missing poor decisions, potentially resulting in lost customers and regulatory fines.
- 6
Security
All systems are vulnerable to attack, and AI is not immune. In addition to cyber threats like theft of secure data or ransomware attacks, firms should be vigilant to identify when bad actors try to gain access to AI systems to manipulate the AI models, which can result in erroneous or fraudulent transactions.
- 7
Systemic Risk
As AI becomes more widespread, it is possible that multiple institutions will adopt the same or similar AI tools, negatively impacting the industry. For example, markets can be adversely affected if numerous firms use the same AI decision to drive markets in one direction.
The future of AI: Addressing ethical concerns
Microsoft has published six recommended areas for the ethical use of AI: Fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability.
In addition, Microsoft, along with Amazon, Google, Meta, and OpenAI, have agreed to the following new safeguards for AI:
- Using watermarking on audio and visual content to help identify content generated by AI.
- Allowing independent experts to try to push models into bad behavior – a process known as “red-teaming”.
- Sharing trust and safety information with the government and other companies.
- Investing in cybersecurity measures.
- Encouraging third parties to uncover security vulnerabilities.
- Reporting societal risks such as inappropriate uses and bias.
- Prioritizing research on AI’s societal risks.
- Using the most cutting-edge AI systems, known as frontier models, to solve society’s greatest problems.
HSO can help
Along with its many benefits, AI poses a number of ethical challenges for financial services institutions. These institutions should work with industry leaders, technical experts, regulators, and other stakeholders to establish recommended guidelines for the ethical use of AI in the financial services industry.
Navigate AI with HSO by your side.
Learn more about how we support Financial Services firms