AI Policy | Chidush Coach

Beta Product Notice

Chidush Coach is in beta. The AI coaching experience is actively being refined. Outputs may be inconsistent, and we are continuously improving accuracy and relevance based on user feedback.

Our Commitments

Chidush Coach uses artificial intelligence to help Jewish nonprofit professionals work smarter. We believe AI should empower people, not replace their judgment. Our approach is informed by established ethical frameworks including the Rome Call for AI Ethics. We use Anthropic's Claude API, which under its commercial terms does not use customer data for model training.

Transparency

We tell you when you're interacting with AI. Chidush Coach is powered by Anthropic's Claude, and we're upfront about what the AI can and cannot do. We don't disguise AI outputs as human expertise.

Privacy & Data Protection

Your conversations are not used to train AI models. We share only the data necessary to generate coaching responses, and we are clear about exactly what is shared. See our Privacy Policy for full details.

Human Accountability

AI generates recommendations; humans are responsible for this platform. Our team reviews AI outputs for quality, monitors for issues, and makes all decisions about how the AI is designed and deployed. If something goes wrong, a person at Chidush is accountable.

Honesty About Limitations

AI can be wrong, outdated, or incomplete. We say so clearly throughout this policy and within the product itself. We will never overstate what AI can deliver.

Fairness & Inclusion

We work to ensure our AI coaching is relevant and useful across the diversity of Jewish organizations — different sizes, denominations, geographies, and missions. We actively monitor for bias in recommendations and correct issues as we find them.

Accessibility

We strive to follow WCAG accessibility guidelines in our design and development because we believe excellent technology should be usable by everyone. We continuously work to improve the accessibility of our platform.

User Control

You decide what to share, what to implement, and when to stop. You can access, export, or delete your data at any time. AI serves you; you are never locked in.

Continuous Improvement

We treat AI ethics as an ongoing practice, not a checkbox. User feedback directly shapes how we refine the coaching experience, and we update our practices as the technology and our understanding evolve.

1. How AI Works in Chidush Coach

Chidush Coach uses Anthropic's Claude, a large language model (LLM), to power its coaching conversations. When you interact with your AI coach, here is what happens:

Your profile information (name, role, organization, tools you use, and professional context shared during onboarding) is sent to the AI along with your conversation messages.
The AI generates responses based on patterns in its training data and the context you provide. It does not have access to the internet, your organization's systems, or any external data sources during the conversation.
Each coaching session type (discovery, building, measurement) uses a specialized set of instructions that guide the AI's behavior and focus.

2. Important Limitations

AI coaching is a powerful tool, but it has real limitations. Please keep the following in mind:

AI Can Be Wrong

The AI may generate inaccurate, incomplete, or outdated information. It can confidently present incorrect facts, suggest tools that don't exist or have changed, or provide time estimates that don't reflect reality. Always verify recommendations independently before acting on them.

Not Professional Advice

AI-generated recommendations are not a substitute for professional advice. Chidush Coach does not provide legal, financial, cybersecurity, compliance, HR, or other professional guidance. If a recommendation touches on these areas, consult a qualified professional before proceeding.

No Guarantee of Results

Time savings estimates, efficiency projections, and impact predictions are approximations based on general patterns. Actual results will vary depending on your organization, technical environment, team capacity, and implementation approach. Past results discussed on the platform do not guarantee future outcomes.

Limited Context

The AI only knows what you tell it. It does not have access to your organization's internal systems, policies, budget, staffing, or technical infrastructure. Recommendations are based solely on the information you share during conversations and onboarding.

Technology Changes Rapidly

The AI's knowledge has a training cutoff date. Tools, platforms, pricing, and best practices evolve quickly. A recommendation that was accurate when generated may become outdated. Always check current pricing, features, and availability of any recommended tools or services.

3. Your Responsibility

As a user of Chidush Coach, you are responsible for:

Evaluating recommendations: Review all AI suggestions critically before implementing them. Consider whether they are appropriate for your organization's specific context, policies, and constraints.
Data sensitivity: Be thoughtful about what information you share in coaching conversations. Avoid sharing financial account details, passwords, or confidential organizational information that you would not want processed by a third-party AI service.
Protecting community members: As a nonprofit professional, you may work with vulnerable populations. Do not share personally identifiable information about program participants, beneficiaries, donors, or community members in coaching conversations. If you need to discuss workflows involving people you serve, use general descriptions or anonymized examples rather than real names or identifying details. All conversation content is processed by a third-party AI provider.
Implementation decisions: You are solely responsible for any changes you make to your workflows, tools, or systems based on AI recommendations. Test changes in a safe environment before deploying them broadly.
Compliance: Ensure that any solutions you implement comply with your organization's policies, applicable laws, and any regulatory requirements relevant to your operations.

4. What Data Is Shared with the AI

To personalize coaching, the following information is included in AI requests:

Your first and last name
Your position/role
Your organization's name and mission
Your key responsibilities
Tools and software you currently use
The full conversation history for the current session
Details of the opportunity/solution being discussed, including problem description and selected solution level (in building and measurement sessions)

Note: Some information collected during onboarding (such as your AI experience level, location, organization size, and estimated hours spent on manual work) is stored in your profile to improve the platform but is not sent to the AI during coaching sessions.

This data is processed by Anthropic via their API. Per Anthropic's commercial API terms, data sent through the API is not used to train their AI models. For full details on what we collect and how it is used, see Section 2 of our Privacy Policy.

5. AI-Generated Opportunity Cards

During discovery sessions, the AI generates “opportunity cards” — structured recommendations with three solution tiers (AI-Powered, Connect Your Tools, and Build Something New). These cards include:

Problem descriptions and proposed solutions
Recommended tools and technologies
Estimated setup time and projected time savings
Difficulty ratings and learning outcomes

All of these are AI-generated estimates and suggestions. They are starting points for exploration, not guarantees. The tools recommended may have their own terms, pricing, and limitations that you should evaluate independently.

6. Future Capabilities

As Chidush Coach evolves, we are exploring ways to expand the platform's capabilities. Areas we may explore in the future include deeper integration with third-party tools, enhanced data analysis and reporting, and collaborative features for teams.

Any new capabilities that meaningfully change how your data is used, processed, or shared with third parties will be clearly communicated before they become available. New features that involve additional data sharing or processing will require your explicit consent. We will update this policy and our Privacy Policy accordingly.

7. Contact Us

If you have questions about how AI is used in Chidush Coach, our ethical commitments, or concerns about any AI-generated content, contact us at:

Chidush

Email: shalom@chidush.org