Current AI Solutions

Data Handling Rules

All staff must follow these rules when using AI tools with client or internal data:

• Never input client confidential data, personally identifiable information (PII), or data covered by NDA into public AI tools (e.g. free-tier ChatGPT, public Claude).
• Approved enterprise tools with data processing agreements may be used with internal data. Check the Current AI Solutions page for approved tools.
• If a client has specific AI restrictions in their contract or MSA, those take precedence over these general guidelines.
• When in doubt, ask the AI working group before proceeding.

Responsible AI Principles

Optis is committed to using AI responsibly across all its operations and client engagements:

• Human oversight: AI outputs are always reviewed by a qualified person before delivery to clients or use in decision-making.
• Transparency: When AI tools contribute materially to a deliverable, we disclose this to clients.
• Fairness: We evaluate AI tools for bias before adopting them, particularly in HR, recruiting, and client-facing applications.
• Accuracy: AI-generated content is fact-checked against authoritative sources. We do not rely on AI for legal, financial, or compliance determinations without expert review.

Approved Tools

The following categories apply to AI tools at Optis:

• Approved for client work: Tools listed in the Solutions page with Active status. These have been reviewed for data handling, licensing, and output quality.
• Approved for internal use only: Tools the working group has vetted for internal productivity but not for client deliverables (e.g. free-tier tools with acceptable data policies for non-confidential work).
• Not approved: Any tool not listed above. Submit an idea through the Idea Submission page to propose adding a new tool.