Q&A on the use of AI in Asset Management

I was interviewed by our CMO recently for my views on the use of AI in Asset Management. Here is how it went:

Question: Hi Aurelien. In your view, to what extent are asset managers considering the use of AI?

Answer: It is definitely early days, but we're seeing a lot of interest. We recently held an event for investor relations teams at GPs where we asked how many were already using AI, how many were actively evaluating and how many were not there yet. Most were investigating or not there yet, but everyone in the room was interested. We also discussed the use cases and anticipated benefits, which included AI enabled targeting and outreach messages during fundraising, using Gen AI for enhanced RFP responses and process automation to streamline operations. These were deemed most likely and most valuable.

In fact, our platform offers a combination of AI and non-AI features that enable our clients to respond to RFPs and DDQs in the most efficient manner. Our clients not only expect a good user experience, but they expect to respond to DDQs as quickly as possible and reduce as much as possible the risk of providing incorrect information. For this to happen, we have carefully looked into empowering our clients to 1) re-use existing content and 2) efficiently create new content.

Question: Thanks, but how exactly are they doing that?

Answer: Dasseti platform acts as a master database, a golden source or the source of truth (SoT) as our clients call it (we call it the “QA bank”). We have developed multiple features that allow our clients to maintain accurate information and to ensure that the questions and responses are always up to date. Our clients typically also store completed DDQs to further enrich it. They would then use this to respond to new questionnaires.

On a given new question or questionnaire, we have developed AI features (“which we call Smart Search”) that will identify and rank similarly asked questions stored in the Dasseti platform. Clients can choose from any of the associated responses to respond to the new questionnaire. In addition, our clients can also leverage our Generative Response feature that will propose an AI generated answer (created using their own content).

The accuracy of these models is remarkable. The key differentiator (when comparing to more traditional searching methods) is that we apply searches based on meaning. This allows our users to easily identify two questions with the same meaning even if these questions have no single word in common.

Question: What are the main concerns that asset managers have around the use of AI?

Answer: The most common concern is Data privacy. In the investment management industry (and many other industries), data confidentiality has always been something that allocators and managers are not willing to compromise. This has always been our top priority, and this was our first consideration when looking into implementing AI for our clients.

Another concern was the impact on roles, with the changing skills requirement being a high priority. IR teams are concerned that if they don't have or gain the skills to manage the systems, they could be replaced by individuals who do have those skills. Of course, the human element is still absolutely imperative in fundraising and IR particularly.

Question: How do you suggest asset managers mitigate concerns?

Answer: At Dasseti, we have mitigated these concerns by:

1. Working with the most established providers that use highest enterprise-grade security standards and that use well established server providers (such as Amazon Web Services)
2. Ensuring that such providers have clear data usage policy that we (and our clients) are comfortable with. For example, our AI provider may securely retain the information for up to 30 days to identify abuse (deleted past that point) and none of the data is used to train or enhance their models.
3. Limiting data sent to such providers - Only relevant data points that are needed to perform the relevant tasks will be sent to the AI provider.

Question: Any there any specific disclosures asset managers should be aware of when using AI tools?

Answer: Managers do not have to disclose the use of AI. Ultimately, managers continue to remain accountable for the accuracy of the data provided. Therefore, it is the managers’ responsibility to ensure that the responses that are provided to clients have been reviewed.

Whether Managers use AI or not, managers will need to continue to ensure that allocators’ data is stored in a secure way. This should be key in developing any AI offering.

Question: Where do you see the biggest gains for asset managers using AI

Answers: There is significant potential for AI to transform RFPs/DDQs processes. Through our deep internal understanding of our asset managers’ process/pain points/vision, we are continuing to develop the Dasseti platform in a way that will empower them to become more efficient, with and without AI. We are continuously assessing the changes within the AI space, and we should only implement new AI features thar really add value to our clients.

Get in touch if you want to find out more about our AI features and roadmap.