Artificial Intelligence (AI)
Acceptable Use Policy

Introduction

ZGF recognizes the transformative potential of Artificial Intelligence (AI) in architectural design. AI technologies have the power to enhance creativity, improve efficiency, and foster innovation. ZGF’s views AI as a tool which may enhance our creative and problem-solving abilities. AI has the potential to augment our design capabilities, but human-centric design principles remain at the core of our projects. As we contemplate how AI may be integrated into our design processes, it is imperative that we establish a clear set of principles to guide responsible and appropriate use. This policy is designed to guide employees in understanding our approach to AI, ensuring that its use aligns with our core values and protects sensitive information.
Questions on the evolving use of AI and AI standards should be forwarded to the ZGF Technology department.

Definitions

IT resources: Any information technology equipment or services that are owned, leased, licensed, or provided by ZGF or accessed through ZGF’s network. IT resources include computers, laptops, tablets, smartphones, printers, scanners, software, applications, networks, internet, email, cloud services, data storage devices, and any other IT equipment or services.
Data Any information or content that is created, stored, transmitted, received, or processed by employees using ZGF’s IT resources. Data includes, but is not limited to, documents, drawings, designs, images, renderings, models, videos, audio files, emails, messages, passwords, personal information of clients or employees, and any other data related to ZGF’s or ZGF’s clients’ business activities or objectives.
AI Artificial intelligence (AI) refers to a computer or software system that can reason, learn, and act in such a way that would normally require human intelligence or that involves data whose scale exceeds what humans can analyze. AI encompasses a set of technologies that are based primarily on machine learning and deep learning, used for data analytics, image analysis, predictions and forecasting, object categorization, natural language processing, recommendations, intelligent data retrieval, and more. Common popular AI tools include ChatGPT, Midjourney, Dall-E, Stable Diffusion, and Copilot.

Policy

Employees are expected to use AI resources and tools in a lawful, ethical, professional, and secure manner that maintains ZGF’s legal and contractual obligations and protects and maintains the firm’s reputation and interests.

Appropriate Use and Practices

ZGF is committed to upholding an ethical set of standards for the use of AI that protects existing copyrights and complies with all applicable laws and regulations, including those governing intellectual property.

• Verify with a project manager before using AI tools on a project. AI use may not be appropriate for every project.
• Employees must restrict the use of AI to ZGF approved AI software. ZGF Technology will manage an evolving list of available options. The Technology department should be contacted with any questions.
• Confirm the validity of any AI output and modify as required to align with industry and ZGF standards. For example, Large Language Models (LLMs) project an air of authority, but may not be relying on valid or accurate references.
• Use AI to inspire creativity and spark ideas.
• Use to accelerate reading, create summaries, or compose first drafts.
• Do not use AI-produced material for anything client-facing.
• Do not use AI for commercial purposes or in a way that directly influences design on a ZGF project.
• Do not present AI-generated content as ZGF work or as a client deliverable.
• Do not use generative image AI tools to replicate specific artists’ original styles or works.

Data Privacy and Security

Uploading information to a public AI system results in public disclosure and strips away confidentiality rights that may have existed before information was uploaded and may violate privacy rights. Protecting the privacy and security of ZGF’s and ZGF’s clients’ Data is critical. Employees must handle ZGF and client Data and other sensitive information with care and in compliance with ZGF’s contractual obligations and applicable data protection laws and regulations. Data's privacy and security must be protected when using AI.

• Do not upload any Data or any specific information regarding the firm, clients, or employees into the public cloud version of a Large Language Model (LLM).
• Never use client or project names in public cloud AI prompts. If necessary, use pseudonyms or placeholder names like, {clientName} or {projectName} then find/replace the placeholders in a revised document.
• Employees must always remove and/or avoid inclusion of ZGF or ZGF client-specific information or Personally Identifiable Information (PII). If you are not sure of the sensitivity of the information, assume it is confidential and proprietary and refrain from inputting into any public AI system.
• Never submit information to any public AI platform without clearly understanding the Terms of Use (TOU). Different AI models have different TOU. These TOU’s detail the terms and conditions governing use, ownership, licensing, etc. applicable to use and content output. Employees are advised to review the TOUs prior to use and to seek legal counsel if needed.
• Check and validate input data before each import to an AI system.
• Treat all public AI input as you would a social media post and assume any information you enter will become publicly available. ZGF’s Social Media Policy in the Employee Handbook applies here.
• Notify your manager of any signs of deliberate or inadvertent misuse of ZGF data or intellectual property or trade secrets.

Transparency

ZGF is committed to being transparent about the use of AI in our design processes. Clients must be informed when AI technologies are utilized on their projects. ZGF assumes responsibility for the accuracy of any AI-driven output.

• Employees are to disclose the use of any AI-generated work to clients.
• All AI results must be verified.
• Each project manager is responsible for the veracity of ZGF Data.

Policy Compliance

All employees are expected to adhere to the provisions outlined above. Employees who do not follow the specified policies may be subject to disciplinary action up to and including termination.

Review and Revision

This policy will be reviewed regularly or as deemed necessary by ZGF’s Design Technology and Information Security teams. Revisions may be made to address any changes in technology, regulations, or business requirements.