AI Responsible Use Policy

CMD is committed to the use of generative artificial intelligence in a responsible manner. Responsible use consists of the following considerations:

1. Privacy – CMD is responsible for any proprietary information our clients provide to us and will ensure that any AI tools used with regard to proprietary info are through a secure and closed framework where the models are not trained on the data that could be output later.
2. Clearance of rights for use – all imagery and copy generated by AI will be vetted for copyright infringement and plagiarism to the best of the agency’s ability. As the generating systems are not always set up to protect artists, there may be limitations in this discovery. CMD cannot guarantee the full clearance of all elements within an AI-generated image or copy block. Where a desired image or script is found to be in violation of copyright law, CMD will proceed as follows:
   1. Research the path to license these assets.
   2. Explore recreating the asset with cleared elements.

   NOTE: Licensing and recreation will likely incur additional costs.

3. Expertise included – all assets generated through AI technologies require an expert touch. At CMD, we ensure that any AI-generated assets are still crafted by experts in the discipline. This process reflects CMD’s commitment to high-quality products as well as the continued employment of our valued experts.
4. Fact-checking and hallucinations – AI models have been known to generate false information or misleading hallucinations. CMD does it’s due diligence in fact-checking any information output by AI but cannot guarantee the accuracy of the output of these models.
5. Bias counteracted – The output of AI often reflects the biases of the data it is trained on. CMD is committed to listening to the diverse voices of our team members and those of the community around us. When prompting and refining, we often request a diverse output from AI. Additionally, our continued education in DEI as a company allows us to better identify and correct bias where it occurs.