Fine Print

Automatic checks · Checked November 16, 2026

Bynder Brand Compliance Agent

Bynder is a DAM (digital asset management) system. Its Brand Compliance Agent checks each asset in the DAM against rules that your admins write from your brand guidelines. The rules can cover the logo, the colors, the typography, the wording, and the claims.

The agent proposes a status for each asset: Compliant, Non Compliant, Partially Compliant, or Not Applicable. A person with the Manage Agents permission approves the status or changes it.

GroupAutomatic checks
Where the work must be
Assets in the Bynder DAM

The agent checks assets that are in the Bynder DAM (digital asset management) system. A user starts a run from the Asset Bank or from a Collection. An admin can set the agent to run automatically when users upload new assets. The admin can limit this trigger by metaproperty, tag, or file type.3

Can an agent run it?
Partly: its API runs it on Bynder assets

The Bynder AI Agents API has a Run agent call. The call takes an agent ID and a list of asset IDs, and it runs the agent on those Bynder assets. Bynder MCP (Model Context Protocol) lets an AI client search assets and search guideline text. Its list of current tools has search, list, and download tools, and no tool that runs an agent.4

File types
Images, PDFs, and office documents

The agent supports PNG, JPEG, GIF, WebP, PDF, PPTx, Doc, and Sheets files. The maximum file size is 25 MB. A PDF can have 30 pages at most.1

What the check gives back
Status and a rule-by-rule report

Each asset gets a proposed status: Compliant, Non Compliant, Partially Compliant, or Not Applicable. A report for each asset shows which rules passed, which rules failed, and the reason. For a PDF, each issue in the compliance report includes the page number.3

Who decides
Proposes, a person approves

The agent proposes a compliance status for each asset. A person with the Manage Agents permission approves the status or overrides it. The decision of that person sets the final status. If an admin turns on Auto-apply, the agent writes its result to the asset without manual approval. An admin can attach permission sets to the values of the compliance metaproperty. Then the visibility of an asset changes when its compliance status changes.3

Where the rules come from
Rules that admins write

DAM (digital asset management) administrators and brand managers write the rules from their brand guidelines and regulations. They group the rules into categories. A rule can have one reference asset, for example a logo file. The agent also follows the Global AI Instructions that the customer sets in the AI Control Center.1

Brands in one account
One agent for each brand

Bynder tells a customer with several brands to set up one agent for each brand. The agent can check against multiple rulesets for brands, regions, channels, or campaigns.1

Plan with the check
Separate purchase, by quote

Brand Compliance is one of the four agent capabilities of the Bynder AI Agentic Platform. The Bynder help center refers to customers who purchase the AI Agentic Platform. After the purchase, a Customer Success contact must enable Automation Workflows for the portal.7

Free planNot published

What does Bynder check?

CheckResultWhat the vendor says
LogoYesThe agent checks logo placement, size, clear space, and color treatment. It also checks logo integrity, which covers distortion, effects, and proportions.1
The vendor's words

Logo placement, size, clear space, and color treatment; logo integrity (distortion, effects, proportions)

ColorYesThe agent takes the dominant colors from an asset. It compares them with the exact RGB, CMYK, or hex values in the rule.1
The vendor's words

Pulls the dominant colors from an asset and compares them against the exact RGB, CMYK, or hex values in your rule

TypeYesIn PDF documents, the agent reads the font family, weight, style, and size from the embedded font data. Typography is also in the Visual identity category of checks.1
The vendor's words

Font family, weight, style, and size, read directly from the document's embedded font data rather than from an image scan.

Copy and termsYesThe Voice and messaging category covers tone of voice, brand voice, and messaging. The agent reads text from image-only or scanned assets, so it can check copy rules on flattened files. The rule guide from Bynder says that brand voice rules check visible text only.1
The vendor's words

Reads text out of image-only or scanned assets, even when the file has no native text layer

Claims and legal linesYesThe Regulatory and legal category covers regulations, market claims, and accessibility. Bynder states that the agent flags non-compliant regulations, claims, offers, technical specs, and forbidden objects.1
The vendor's words

flags non-compliant regulations specified by you, claims, offers, technical specs and forbidden objects

The fine print

Limits that Bynder states

  1. Very small or low-confidence text (under roughly 0.15% of the image area) may be skipped by text detection

    Text detection can skip very small or low-confidence text, under about 0.15% of the image area. Rules that depend on fine print or small disclaimers can give unreliable results.1

  2. Color matching runs on a downscaled copy of the image, so very fine color detail may not be captured at full precision.

    Color matching runs on a smaller copy of the image. The agent can miss very fine color detail.1

  3. Font rules are marked as not applicable rather than pass or fail.

    If a PDF has no text that the agent can extract, the agent marks the font rules as not applicable.1

  4. Maximum 1,000 assets per run.

    One run can check 1,000 assets at most.3

  5. Maximum 50 rules or 10K characters per agent, across all categories combined.

    One agent can have 50 rules or 10,000 characters at most, across all categories.1

  6. The agent checks one asset at a time, not the full campaign package.

    The agent checks one asset at a time. It cannot check a rule about a full campaign or a set of assets.2

  7. Avoid rules that require information not visible in the asset and not available as a file property

    The agent cannot check a rule that needs information that is not visible in the asset and is not a file property. Examples are license status and approval history.2

  8. AI can make mistakes. We recommend that you only apply automated compliance evaluations after careful consideration

    Bynder states that AI can make mistakes. Bynder tells customers to apply automated compliance results only after careful consideration, and to consult teams such as legal counsel.3

Which plans does Bynder have?

Bynder publishes no prices. Bynder customizes each plan and gives the price in a quote.9

Prices are from the vendor's pricing page.

What else do Bynder's pages say?

Overrides are logged
Bynder logs each compliance decision that a person changes in the History tab of the agent. The original AI report stays on the detail page of the asset.3
Agent APIs for external systems
Bynder states that its Agent APIs make the Enrichment Agent and the Brand Compliance Agent callable from an external system.10
Rule tests are free
A user can preview and test rules on assets from the DAM (digital asset management) system before the agent goes live. Bynder states that setup and testing are free of charge.3
Two permissions
A user needs the Manage Agents permission to build rulesets and to approve proposed statuses. A user needs the Run Agents permission to run checks on assets.3
Can check file properties
Rules can also check the file properties of an asset. Examples are resolution, dimensions, file format, file size, tags, copyright, dates, and archive status.1
Photography, format, and safety checks
The agent also has check categories for photography and imagery, channel specifications, and safety. The safety category covers personal protective equipment in images for industrial customers.1
No training on customer data
Bynder states that it does not use customer data to train AI models. Bynder names Anthropic (through AWS Bedrock), Google, and AWS as the AI providers for its AI Agents.10

Who does it suit?

It suits a team whose finished files already go into Bynder. An admin can set the agent to run when users upload new assets, so each file gets a status before people share it. The status can also change who can see the asset.

If your code must fetch the files, The Brand API Register describes the Bynder API.

When is another product a better fit?

  • If agents make your work outside Bynder, the work must go into the DAM before the agent can check it. The AI Agents API starts a run only on assets that are in Bynder. Rainbrand and Markup AI check work that an agent sends to them.
  • If your rules depend on small legal text, test them first. Bynder says that text detection can skip very small text.
  • If named people must comment on each file and sign it off, use a proofing tool such as Filestage or Ziflow.

Compare

Where are the vendor's own pages?

Sources

  1. support.bynder.com: Brand Compliance Agent Overview and Capabilities checked November 16, 2026
  2. support.bynder.com: How to Write Good Brand Compliance Rules checked November 16, 2026
  3. support.bynder.com: How to Configure and Use the Brand Compliance Agent checked November 16, 2026
  4. api.bynder.com: runagentexecution checked November 16, 2026
  5. api.bynder.com: mcp checked November 16, 2026
  6. support.bynder.com: Q3 2 Release Notes checked November 16, 2026
  7. support.bynder.com: AI Agentic Platform Product Overview checked November 16, 2026
  8. support.bynder.com: How to Use Bynder's AI Automation Workflows checked November 16, 2026
  9. bynder.com: pricing checked November 16, 2026
  10. bynder.com: ai agents checked November 16, 2026