CasaPuente • June 2026
Staff AI Survey Results
Overview
AI Adoption & Usage
Functional Areas
Tools & Tasks
Concerns & Risks
Opportunities
Training & Support
Key Takeaways
Respondent Detail
CasaPuente · The Big Picture
Staff are actively experimenting with AI, with strong appetite for guidance, training, and organizational clarity.
52
Respondents
Staff AI Survey, June 2026
Active Users
Daily or weekly at work
Top Concern
Training Demand
Selected at least one topic
How to navigate: Click any tab above to open one section in detail, or stay here in the Overview and expand the cards.
AI Adoption & Usage
Nearly two-thirds of staff are experimenting or actively using AI. A tenth have made a principled choice to opt out.
Adoption is uneven. Experimenters lead at 31%, with Regular and Integrated Users combining for 40%. One in ten staff are intentional non-users — their perspective should inform, not be overridden by, CasaPuente's AI strategy.
61%
Regular+ Users
Experimenter, Regular & Integrated
User Archetype & Approach
Q4Which best describes your current approach to generative AI in your work?
▶Staff Voices
Usage Frequency
Q2How often do you use a generative AI tool like ChatGPT, Copilot, Gemini, or Claude? (Select one per column)
Reasons for Not Using AI at Work
Q6If you rarely or never use generative AI for work, what is the primary reason? (Select one)
▶Staff Voices
Functional Areas
Adoption varies sharply by department.
Executive, HR & Operations leads on active use (44%). Programs & Community Impact is the most divided, with a strong intentional-non-user population (40%) alongside active users. Finance is the largest team (17 staff) but the most unadopted (6% active), making it the highest-leverage early focus.
5
Functional Areas
Finance leads by headcount (17)
User Archetype by Functional Area
Q1 + Q4Functional area cross-tabulated with current AI approach
Tools & Tasks
Built-in AI features lead at work, consumer AI tools dominate personal use.
Five different AI tools are in active workplace use with no coordinated approach, creating an unintentional data-privacy exposure pattern. Of these, only Google Gemini and Microsoft Copilot Chat are sanctioned/licensed; ChatGPT and Claude run on individual accounts staff set up themselves (personal or CasaPuente email), which admins cannot distinguish from the console — the admin visibility gap. Writing & communication and operations are the dominant tasks, pointing to clear opportunities for shared workflows. Note: this survey counts the five standalone chatbots staff named (ChatGPT, Copilot, Claude, Gemini, and similar); the Assessment Summary’s “seven” additionally counts embedded AI in already-owned tools (Power BI AI, Power Automate, IBM SPSS).
46%
Use Built-in AI
Most-used tool at work
Tools Used
Q3In the last month, which generative AI tools have you used? (Check all that apply, at work and/or personally)
Task Types
Q5If you've used generative AI at work, what types of tasks have you used it for? (Check all that apply)
Concerns & Risks
Environmental cost and data privacy lead the broad concerns (tied at 55%). The role of human judgment dominates equity-specific concerns.
Staff answered two related questions: a broad read on values-level concerns about AI at CasaPuente, then a focused read on AI's equity implications. The two views are complementary, not redundant. Both confirm that concerns aren't fringe; they're shared across departments and archetypes, including among CasaPuente's most active AI users.
44%
Top Mission Concern
Data privacy / misuse
Broad Concerns about AI Use
Q7What mission- and values-related concerns do you have about CasaPuente using AI? (Select up to three)
▶Staff Voices
Concerns about AI & Equity Specifically
Q8When you think about AI and equity in our work, what concerns you most? (Select up to two)
▶Staff Voices
Opportunities
Freeing staff time and reducing reporting burden lead the value opportunities.
69% of staff see the top value as freeing up time for higher-value work. Translation & accessibility and reducing grantee/client reporting burden tie as strong mission-aligned opportunities (54% each). AI is framed as a capacity multiplier across the organization.
69%
See Productivity Value
Free staff time for higher-value work
AI Use Case Opportunities
Q9Thinking about the work we do, in which areas do you see the biggest AI use case opportunities? (Select up to three)
▶Staff Voices
Value Creation Opportunities
Q10And thinking about the kind of value AI could create, where do you see the biggest opportunities for CasaPuente? (Select up to three)
▶Staff Voices
Training & Support
Staff want training across every topic. Evaluating outputs, role-specific discovery, and data privacy lead the list.
Training appetite is broad across the organization. Evaluating AI outputs for accuracy and bias (54%) and role-specific AI discovery (52%) lead the list, closely followed by data privacy (50%). Building AI-assisted workflows (38%) and AI ethics & equity (31%) round out the top five.
28
Want Output Eval Training
Top training topic
Training Topics
Q11Which AI training topics would be most beneficial for your day-to-day work? (Select your top three)
▶Staff Voices
Support Format Preferences
Q12Which formats of support would be most helpful for you? (Select your top two)
Key Takeaways
Four lenses for moving forward.
Ethical, governance, strategic, and training implications synthesize the findings into a path forward, with current-state evidence and future-state opportunities for each.
4
Lenses
Ethical · Governance · Strategic · Training
Ethical
Current State
Data privacy and role diminishment lead mission concerns.44% cite data privacy / misuse and 31% cite role diminishment. Staff are concerned about what happens to sensitive client and community data when entered into AI tools.
Loss of human judgment and skill atrophy tie as follow-on concerns.Each named by 31% of respondents. Staff worry most about AI displacing the human relationships, empathy, and lived-experience judgment central to CasaPuente's mission.
10% are intentional non-users.A principled position concentrated in roles most grounded in community trust. Their stance reflects organizational values that should inform AI strategy, not be overridden by it.
Future State Opportunities
Define data classification tiers for AI use.Be explicit about what client, community, and organizational data may and may not be shared with AI tools. Staff need a clear, practical reference — not just a general policy statement.
Define which decisions stay human-led.Especially in direct service, intake, and community-facing work, where staff have expressed clear discomfort with AI involvement. Make these boundaries explicit policy, not left to individual discretion.
Engage intentional non-users as ethical advisors.Their principled stance reflects organizational values that should inform AI strategy, not be overridden by it. Include them in governance design conversations.
▶Staff Voices
Governance
Current State
50% are experimenting informally or waiting to start.31% Experimenters and 19% Curious non-users. The absence of official guidance has not stopped experimentation — it has made it unsanctioned and uneven across the organization.
Tool use at work is fragmented.Built-in AI 46%, ChatGPT 40%, Copilot 37%, Claude 31%, all in active use with no coordinated approach. The result is inconsistent and untracked data privacy exposure across the organization.
Data privacy is the 3rd most-requested training topic (50%).Staff explicitly ask for guidance on what is safe to share in prompts, how internal and client data will be protected from AI providers, and what expectations apply to program participants.
Future State Opportunities
Publish a practical AI use policy with tiered data sensitivity.What is safe to share, what requires caution, and what is off-limits. Extend the policy to partner-facing and client-facing use as well as internal operations.
Standardize on a small number of sanctioned tools.With five tools already in workplace use, CasaPuente needs to rationalize its AI stack, establish licensed agreements where appropriate, and give staff clear guidance on which tools to use for which purposes.
Build governance with broad staff input.Staff need inclusive process, not leadership-down policy. Include frontline and community-facing staff whose roles carry the highest relational stakes in any AI deployment decisions.
▶Staff Voices
Strategic
Current State
40% are active or integrated AI users.21 respondents (Regular + Integrated), spread across functional areas. CasaPuente has meaningful early adopter energy that can anchor peer learning — if channeled with organizational support.
Freeing staff time is the dominant value opportunity (69%).Followed by reducing grantee/client reporting burden (54%) and translation & accessibility (54%). Staff see AI as a capacity multiplier, not a replacement for judgment.
Finance / Investments / Admin is the largest team with structured workflow potential.17 staff in finance/admin represent the largest single functional area. Data analysis, reporting, and workflow automation are high-leverage, lower-relational-risk starting points.
Future State Opportunities
Invest in translation and accessibility as a mission-aligned priority.54% cite translation & accessibility as a top value opportunity — one of the highest areas of both strategic and mission alignment for CasaPuente's community-facing work.
Reduce grantee and client reporting burden.54% see this as a top value opportunity. AI-assisted summarization, intake support, and plain-language translation can meaningfully lower barriers for the communities CasaPuente serves.
Start with lower-stakes, high-volume workflows.Research synthesis, report drafting, and writing support — where Research & Analysis (67%) and Writing & Communication (62%) are already the most common tasks — build confidence before moving into more sensitive territory.
▶Staff Voices
Training & Development
Current State
Evaluating outputs and bias is the top training need (54%).Named by more than half of respondents. With staff actively using a range of tools, the need to critically assess AI-generated content is immediate and practical — not theoretical.
Role-specific AI discovery is the 2nd priority (52%).Staff want to understand how AI can concretely help in their own jobs, not receive generic training. This points to the value of department-level or role-based workshops over all-staff sessions alone.
Data privacy and AI ethics each rank in the top five.Data privacy (50%) and AI ethics & equity (31%) both reflect CasaPuente's mission-driven culture. Staff want to use AI responsibly and need clear organizational guidance to do so.
Future State Opportunities
Lead with output evaluation and data privacy as universal first modules.Both rank in the top three across the staff population. They are the highest-impact, lowest-risk starting points — building critical thinking skills and trust before expanding adoption.
Build role-specific learning paths.Pair an all-staff foundation module (data privacy, ethics, limitations) with department-level sessions that show how AI concretely helps each team's work. Avoid one-size-fits-all delivery.
Prioritize workflow-building for the Experimenter cohort.With 31% of staff already experimenting, there is a ready cohort for applied workflow-building training (38% requested it). Systematizing their ad hoc use creates reusable templates the whole organization benefits from.
▶Staff Voices
Respondent Detail
Every respondent, every concern, every comment.
Sort, filter, and click any row to isolate the respondent across the whole dashboard. Open comments (Q13) are organized by which question they relate to.
52
Respondents
Click any row to isolate
Respondent Data
All Respondents
Open Comments (Q13)
Filtering by:
| Respondent | Functional Area ↕ | AI Approach ↕ | Freq. Work ↕ | Tools (Work) | Mission / Values Concerns (Q7) |
|---|
Q13Is there anything else we should consider as we begin to develop our AI governance, training, and deployment strategies?
The work featured in this example involves CasaPuente, a fictional entity created solely for demonstration and portfolio purposes. Any resemblance to actual organizations, real-world brands, or existing entities is purely coincidental.
About this Survey. Staff were invited to take a 13-question survey on generative AI: how they use it (or don't), what concerns and opportunities they see, what tools they reach for at work and personally, and what training and support they want. The intent of the survey is to ground CasaPuente's AI governance, training, and deployment strategies in actual staff experience and values rather than assumptions.
About this Survey. Staff were invited to take a 13-question survey on generative AI: how they use it (or don't), what concerns and opportunities they see, what tools they reach for at work and personally, and what training and support they want. The intent of the survey is to ground CasaPuente's AI governance, training, and deployment strategies in actual staff experience and values rather than assumptions.
Confidential | Generated by beneAI | CasaPuente Staff AI Survey · June 2026