Expert: Catherine Parry, Deepview. Facilitator: Niall Buggy, Sionic
The role of transparency, regulation, and definition of ‘artificial intelligence’ (AI) with respect to Wealth Management and Advice needs careful consideration.
- For Advisers, there are many use cases that AI can be applied to such as supporting advisers in data analysis, data cleansing and case preparation.
- Current data governances and architectures could be stretched by future AI use cases and Client privacy requirements.
- Operational Resilience and Security are operational areas where AI can provide support for Wealth Managers.
Key Challenges, Conclusions and Solutions
Artificial intelligence is revolutionising how consumers and companies alike access and manage their finances. The global AI fintech market is predicted to reach $22.6B in 2025.
In fact, 70% of all financial services firms are using machine learning to predict cash flow events, fine- tune credit scores and detect fraud, according to a survey by Deloitte Insights.
An Economist Intelligence Unit (EIU) research report found that 86% of Financial Services executives plan on increasing their AI-related investments through 2025. The study analysed the sentiments of 200 business executives and C-suite leaders at investment banks, retail banks and insurance companies in North America, Europe and Asia-Pacific.
1. Regulation of AI in Wealth Management
The role of transparency, regulation, and definition of “artificial intelligence” (AI) with respect to Wealth Management and Advice needs careful consideration.
The term AI is used as an all-encompassing buzzword for machine learning, pattern analysis, and algorithms (a non-intelligent set of rules or process). There needs to be distinction between automating a human process and “real AI”, where humans may not be able to explain how the outcome is reached.
This raises the question of transparency, and where do we draw the line between advice and the suggestions of a computer model. Should an adviser always be able to explain to the client how an advice outcome was reached?
Pattern analysis of client cases and outcomes is a powerful tool to check the Adviser’s analysis as a second line of defence, but when does the review become the output? What role would reputational pressures play?
2. AI Uses Case for Advisers
For Advisers, there are many use cases that AI can be applied to such as supporting advisers in data analysis, data cleansing and case preparation.
Paper and paper equivalent technology are still an important topic in many firms; however, machine learning can be deployed in several use cases today.
The first is to review the consistency of client outcomes and advice provided. Machine learning can be used to analyse cases and identify patterns or data points that are not apparent to the Adviser.
The second use case is being a virtual assistant to the Adviser to perform quality checks and “heavy lifting” tasks in the preparation of advice cases such as compliance analysis.
A third use case is to review documentation and old fact finds. Many notes are written/scribbled on application forms (front and back), and an AI system could read and organise these unstructured notes and analyse historical notes whether computer, or paper based.
In the future, it is likely there will be models to predict changing risk appetites and lifestyle goals, even before the client is aware of them. These models will help drive the conversations with the client for the Adviser.
There are risks of increasing levels of predictive analysis such as client churn, fear of the solution, and the relative capabilities of smaller organisations.
3. Challenges to data governance and architecture
Current data governances and architectures could be stretched by future AI use cases and Client privacy requirements.
Data governance and architecture are strategic imperatives, and the use of AI will emphasise this drive.
The data requirements of AI models and their subsequent outputs may challenge our current understanding of what is, and is not, client data and consent.
The role of the Client Lifecycle Management systems with AI will become more important than CRM as the AI drives the engagement model.
In the fund selection/due diligence process, how would the personality traits of Fund Managers be captured and analysed.
4. Operational Resilience and Security
Operational Resilience and Security are operational areas where AI can provide support for Wealth Managers.
Advances in AI technologies, such as those of DeepView, are highlighting emerging security and resilience risks such as the use of social media and messaging systems such as WhatsApp.
Hackers and social engineers can use social media platforms to build profiles of organisations and their staff through the analysis of posts and photographs. Being able to analyse these risks are critical to corporate security.
Furthermore, identifying work and non-work conversations on messaging apps using AI, ensures corporate compliance and staff privacy without impacting the ease of conversation channels.