Open Forum

We are working on ours as well, but in general are looking at something simple like:

(Institution) does not use generative AI. All work produced by (Institution) is the product of research and writing completed by staff members and/or volunteers utilizing archives and historical records.

Internally we are ok with things like talk to text for Oral Histories (as reviewed by both the interviewer and interviewee) and notes summary tools for meetings.

I'm interested to hear from others!

We are working on ours as well, but in general are looking at something simple like:

(Institution) does not use generative AI. All work produced by (Institution) is the product of research and writing completed by staff members and/or volunteers utilizing archives and historical records.

Internally we are ok with things like talk to text for Oral Histories (as reviewed by both the interviewer and interviewee) and notes summary tools for meetings.

I'm interested to hear from others!

Hi Jill,

We have just finished writing an AI policy. After many revisions, here it is:

Artificial Intelligence (AI) Use

·         SHS/SHM recognizes that AI tools are embedded in the software staff uses every day.  As such, AI tools may be used to support, but not to replace, appropriate human review, judgement, and approval.  Users must follow any licensing, privacy, security, copyright, and recordkeeping requirements that apply to their use of AI tools.

·         Users must not enter, upload, paste, or share confidential information, personal information, payment card information, passwords, membership lists, donor records, employee records, or other sensitive SHS/SHM information into public or unapproved AI tools.

·         Use of AI to draft, summarize, analyze, edit, or produce SHS/SHM work product must be disclosed to the appropriate supervisor, Executive Director, or other designated reviewer before the work product is finalized, shared, published, or relied upon.  Such use must also be disclosed on any final product in which AI played a role in developing, creating, or producing.

·         AI-generated content must be reviewed for accuracy, appropriateness, bias, privacy concerns, and consistency with SHS/SHM policies before it is used, shared, published, or relied upon.

·         AI tools must not be used in curation, creation of exhibits, researching or writing of historical articles, or other similar final products

·         AI tools must not be used to make final decisions about employment, volunteers, members, donors, customers, vendors, security access, disciplinary action, or other matters requiring authorized leadership review.

Suspected misuse of AI tools, disclosure of sensitive information through AI, or inaccurate AI-generated information that may affect SHS/SHM operations must be reported promptly to the designated administrator or Executive Director.

This is where we're sitting with it for now- I'm sure there will be further tweaks to it, but needed to get the basics covered for Board approval this month.

Would love to see what everyone else has; it's such a complex, quickly changing subject!