Responsible AI Policy

Elevate Small Business is committed to the responsible development and deployment of artificial intelligence features across our platform. This policy describes how we govern AI, what safeguards we maintain, and what responsibilities you hold as a user of AI-assisted features.

1. Our AI Principles

• Transparency — We disclose when AI is used and what models power each feature. AI-generated outputs are labeled as such.
• Human Oversight — AI outputs are advisory. No consequential decision is made without human review and approval.
• Fairness — We monitor AI outputs for bias and take corrective action when patterns of unfair treatment are identified.
• Data Minimization — We send only the data necessary for each AI task. We do not train models on your data.
• Accountability — We maintain audit logs of AI usage and take responsibility for the governance of features we deploy.

2. AI Features and Models

The platform uses AI for the following task categories:

• Extraction — Pulling structured data from unstructured documents (financial statements, contracts, filings)
• Interpretation — Categorizing accounts, classifying transactions, and identifying patterns in financial data
• Narrative — Generating written analysis such as investment memos, executive summaries, and business plan sections
• Synthesis — Combining multiple data sources into consolidated insights and recommendations
• Creative — Drafting communications, marketing content, and document templates
• Assembly — Composing structured documents from component data and templates

These features are powered by Google (Gemini) and Anthropic (Claude) models. See our Data Processing Agreement for the full sub-processor list.

3. Prohibited Uses

The following uses of platform AI features are prohibited:

• Making fully automated lending, credit, or investment decisions without human review
• Using AI outputs to discriminate against individuals or groups based on protected characteristics
• Representing AI-generated content as human-authored analysis in contexts where the distinction is material (e.g., regulatory filings, sworn statements)
• Using platform AI features for surveillance, tracking, or profiling of individuals beyond the scope of legitimate business due diligence
• Intentionally circumventing safety filters or prompting models to produce harmful, deceptive, or illegal content
• Using AI features to generate intentional disinformation or to deceive third parties

4. Human Oversight for Consequential Decisions

The platform assists with decisions that have significant impact on individuals and businesses, including capital allocation, lending eligibility, procurement matching, and due diligence screening. For these consequential decisions:

• All AI outputs are advisory only. The platform presents AI-generated analysis as a starting point for human judgment, not as a final determination.
• No loan approval, denial, investment recommendation, or procurement award is made solely by an AI system without qualified human review.
• Users are responsible for reviewing AI-assisted outputs before acting on them in any capacity that affects rights, opportunities, or financial outcomes for others.
• Where AI contributes to a scored or ranked output (matching, risk assessment, eligibility screening), the scoring methodology is documented and reviewable.

5. Accuracy and Limitations

AI-generated outputs are probabilistic. They may contain errors, inaccuracies, omissions, or reflect biases present in training data. Specifically:

• AI-generated financial analysis does not constitute professional financial, legal, or tax advice
• Account categorizations and transaction classifications should be reviewed for accuracy before use in reporting or compliance
• Generated narratives (investment memos, business plans) are drafts requiring professional review and editing
• Data extracted from documents should be spot-checked against source materials

We do not guarantee the accuracy, completeness, or fitness for purpose of any AI-generated output. See Section 6 of our End-User License Agreement for full limitation of liability terms.

6. Data Handling for AI Processing

• Data sent to AI models is encrypted in transit and processed only for the requested task
• Our AI providers (Google and Anthropic) do not use your data to train their general-purpose models
• We apply data minimization: only the fields necessary for each specific AI task are transmitted to the model provider
• Personally identifiable information is redacted or anonymized where feasible before AI processing

For complete details on data processing, sub-processors, and your rights under GDPR and CCPA, see our Privacy Policy and Data Processing Agreement.

7. Transparency and Disclosure

We maintain transparency about AI usage through the following mechanisms:

• AI-assisted outputs are visually indicated in the user interface so users can distinguish them from human-authored content
• Platform administrators have access to AI usage audit logs showing which features were used, when, and by whom
• This policy is publicly available and versioned. Changes are communicated in advance.

8. Feedback and Correction

If you encounter an AI output that is inaccurate, biased, or inappropriate:

• Use the in-platform feedback mechanism to flag the specific output
• Contact us at support@elevatesb.com for concerns about systemic AI behavior

We review flagged outputs to identify patterns and improve the quality and safety of our AI features.

9. AI Audit and Monitoring

Elevate Small Business maintains a comprehensive audit trail of AI feature usage. Each AI invocation records:

• The feature and task type
• The model used
• Token usage and processing duration
• Success or failure status
• The requesting user

This audit data is available to platform administrators through the unified audit timeline and is retained in accordance with our data retention policies.

10. Your Compliance Responsibility

You and your organization are responsible for:

• All decisions made, advice given, and actions taken based on AI-assisted outputs from the platform
• Ensuring your use of AI features complies with applicable laws, rules, and regulations, including laws specific to artificial intelligence (such as the EU AI Act)
• Implementing appropriate human oversight and review processes for consequential decisions informed by AI outputs
• Evaluating AI-generated outputs for accuracy and appropriateness before relying on them

11. Changes to This Policy

We may update this Responsible AI Policy as our AI capabilities evolve, as regulations change, or as we identify improvements to our governance practices. Material changes will be communicated through platform notifications and noted on this page with an updated effective date.

Contact

For questions about this policy or our AI governance practices:
support@elevatesb.com