AI – survey results and where we go from here

June 2024
We have already gotten survey responses from half of our member firms; other firms have indicated that they are in the process of collecting the information. We encourage firms to keep the responses coming. The first survey result: it is clear that this topic is important to our members!


Survey respondents have taken care to gather input from various departments within their firm, forwarding the survey to their firm’s AI task force, etc. to ensure that we received responses based on firms’ overall AI priorities. Many firms noted that all of the aspects were important to their firm, but to different degrees for different department roles.

Firms also added details in the comments section that provided additional focus to topics they highlighted as priorities. Participants from member firms can request the slide deck from the June meeting for details and next steps. To summarize, we have learned that:

  • The topics deemed of highest priority were fair use and disclosure.
  • Member firms recommend that RIXML act as a facilitator of discussion among member firms, leveraging its long history of bringing vendors, sellside firms, and vendors together to discuss these types of issues.
  • Member firms would also like RIXML to act as a forum for education, both by bringing in outside experts and by providing a forum for member firms to share their expertise.
  • In the arena of content creation, summaries created by third parties and content consumers have particular considerations:
    • who is responsible for liability and accuracy when third parties or content consumers create synopses/summaries of sellside research reports?
    • are entitlements and distribution authorization the same as or different than for the full research content they describe?
    • what are the RIXML tagging implications?

At the June meeting, we received a suggestion to open the AI survey up to a broader audience, not just member firms. If you have any suggestions regarding this - firms to include, additional questions to ask, or other ideas for gauging AI-related priorities in our industry - This email address is being protected from spambots. You need JavaScript enabled to view it..


During the discussion of the fair use aspects of AI, one member asked whether the RIXML Research Standard has a mechanism to flag certain content within a research report as being restricted from being fed into AI tools; for example, third-party content that a sellside firm has permission to use within their research reports but that is restricted from other uses. While the current version of the RIXML Research Standard does not have such capabilities, the upcoming release does have significantly enhanced capabilities to tag individual components within a research item. It would be extremely helpful to get input from member firms so that we can better understand what types of content might need to be identified, what information might need to be conveyed, etc. This will help us determine whether our current planned capabilities will provide the needed tags; if not, we will determine whether including additional tags in the upcoming release makes sense or if it would need to be part of the planning for v3.1.
 
As we determine how RIXML can best help member firms address the above, we want to make sure that we:

  • ensure that we are not duplicating efforts underway elsewhere
  • determine how best to support RIXML members and the industry – extending the survey may be a way; opportunities to learn, including about regulatory issues
  • identify any critical AI-related needs for upcoming XML standards release
  • determine action plan and timeline for each key issue based on its needs. At present, fair use seems to be the most pressing issue for our members.

 
We will be convening a meeting to further discuss the findings of the AI survey, determine how best to gather additional input, and begin addressing some of the pressing issues we have identified. We will send the meeting information to the whole RIXML distribution list, but feel free to let Deirdre know if you would like to be involved.