AI creative guidance
Overview
In an AI-powered world, it’s easier than ever to create videos, visuals and copy almost instantly. But as content creation accelerates, our role as brand stewards becomes even more important.
At Kyndryl, we have a responsibility to ensure consistency, clarity and integrity across everything we create, helping us to build trust while mitigating legal and reputational risk. Our brand standards for AI usage provide the governance and guardrails needed to ensure that everything we produce continues to reflect who we are as a brand.
General guidance
The following standards apply to AI use across internal and external brand touchpoints, including advertising, videos, photography, visual systems, and other creative materials.
01
Use only Kyndryl-approved AI tools and practices.
Only use tools and practices approved by Kyndryl AI Governance. This helps to ensure that we’re using AI responsibly and helps protect us from legal, security and data risks.
02
Keep AI-generated images on brand.
All AI-generated content must align with Kyndryl’s visual identity standards to ensure that everything we create continues to reflect who we are as one unified Kyndryl.
03
Use only licensed reference content.
Kyndryl must own the rights to any content used as a reference or input for AI. This helps protect us from legal and data risk.
04
Do not use AI-generated people or avatars
AI-generated humans, avatars, or synthetic representations are not permitted in any branded content. All people depicted must be real individuals with approved usage rights.
05
Do not alter core brand identity of Kyndryl or Kyndryl created images or assets using AI
Branded assets—any image used in advertising, social or content and that has a logo on it, must never be modified, recreated or extended using AI to alter core brand identity (e.g., logos or visual systems). However, generative AI utilities are permitted for technical formatting, canvas extensions and standard technical retouching of approved photography and video assets, provided the original asset's intent and perspective are preserved.
06
Generative AI can be used for the following approved production workflows:
Generative AI utilities are permitted strictly for technical optimization, cleanup and formatting adjustments of human-captured content. This includes removing temporary background distractions, adjusting wardrobe fit or fabric wrinkles, and expanding non-descript background environments (such as skies, textures or repeating architectural lines) to adapt an asset to multi-channel aspect ratios. AI tools must never be used to fabricate entirely fictional events, settings or narrative truths.
07
Any visual asset modified, extended or optimized using generative AI must be reviewed by a Brand Creative Lead.
Any visual asset modified, extended or optimized using generative AI utilities must undergo a final “Human-in-the-Loop” quality review by an authorized Brand Creative Lead via a Workfront request. This review is mandatory to verify visual integrity, ensure perspective and geometric accuracy, and eliminate technical artifacts or algorithmic distortion before the asset is published to internal or external channels.
AI misuse
Avoid or do not use AI in ways that compromise quality, accuracy, or ownership. This includes:
Avoid using generative AI to produce uncontrolled or unverified text.
Avoid visuals with artifacts, or distortions that alter and reduced image quality.
Avoid imagery with inaccurate or unrealistic human anatomy.
Do not use AI to edit or remove watermarks.
Do not use AI to retouch or modify our corporate advertising assets to avoid copyright and legal issues.
Do not create or modify icons, logos or wordmarks.