How many of your failed AI initiatives were “good failures” (i.e., resources used were worth the learnings/innovation)
All/almost all10%
Most65%
Some21%
A few4%
None
80 PARTICIPANTS
All/almost all10%
Most65%
Some21%
A few4%
None
Never6%
Sometimes70%
About half the time17%
Most of the time4%
Always
It is safe to use OpenAI API28%
Not Safe to use OpenAI API45%
Not Sure if safe or not34%
Yes we are confortable building and training our own LLMs14%
No we are not confortable building and training our own LLMs5%
No selling.
No recruiting.
No self promotion.
Rules of EngagementFAQsPrivacy
© 2025 Gartner, Inc. and/or its affiliates. All rights reserved.