API changes in the April 2026 release

The April 2026 release of Adobe Learning Manager introduces focused enhancements to the Public API around alternates and equivalents, time‑windowed access to content, content‑driven quiz attempts, non‑logged‑in experiences, and Job Aid handling. The changes are designed to be largely backward‑compatible while enabling more precise integrations.

Adaptive learning for Learning Paths

This release adds full Public API support for adaptive learning paths. Learning Paths (LPs) can now define adaptive rules that control which sections and sub‑learning objects (sub LOs) are visible to different learner groups, and the Public API reflects this behavior.

The following endpoints are affected:

A new Boolean attribute, attributes.isAdaptive, indicates that a learning program uses adaptive rules. When this flag is true, the sections attribute is interpreted adaptively.

For learner calls, only sections that are visible to the current learner are returned. Each section includes the list of learning object IDs (loIds), a mandatory flag and a mandatoryLOCount that are computed based on the adaptive configuration for that learner, as well as the sectionId. The relationships.subLOs relation is now also filtered, so it contains only those sub‑learning objects that are visible to that learner.

For admin calls, sections can additionally expose an adaptiveConfig array. This describes the adaptive rules per user group, including userGroupId, userGroupName and whether the section is mandatory for that group. Admin‑facing tools can use this to visualize and manage adaptive rules.

Reset completion for learning programs

The release introduces an API to refresh (reset) completion for learning objects, including adaptive learning programs. This supports scenarios such as re‑training, periodic compliance refreshes, or program restarts.

A new endpoint is available:

POST /primeapi/v2/learningObjects/{loId}/instances/{loInstanceId}/refreshCompletion

This operation requires appropriate write permissions and resets the completion state for the specified learning object instance in the given user context. It is intended for use by admin‑side automations or carefully scoped learner tools.

A bulk version of this capability is planned via a Jobs API. The request shape is designed to target users by email, user ID or user group ID, for example:

{
  "emails": ["[user1@example.com](mailto:user1@example.com)", "[user2@example.com](mailto:user2@example.com)"],
  "userIds": ["12345", "67890"],
  "userGroupIds": ["ug1", "ug2"]
}

Integrations should use this API when they need to restart learners in each program or course. Clients must handle error responses gracefully: the API may reject refresh requests, where a reset isn’t applicable (for example, when no completion exists or unsupported learning object types).

Equivalents and alternate completions

To support implementations where multiple learning objects can satisfy the same requirement, the release introduces Equivalents and Alternates completions as public APIs.

The Learning Object endpoints now contain this information:

- GET /primeapi/v2/learningObjects
- GET /primeapi/v2/learningObjects/{loId}

A new Boolean attribute attributes.isAlternateComplete indicates whether the learner’s completion for a given learning object is the result of an alternate or equivalent learning object rather than the object itself. When this is true, the relationships.alternateCompletions relation lists the learning objects that acted as alternates. This allows downstream reporting and dashboards to distinguish between direct and alternate completions and to show which alternate fulfilled the requirement.

In addition, a related‑learning‑objects view allows discovery of potential alternates that can satisfy a learning object. This is exposed via:

GET /primeapi/v2/learningObjects/{loId}/relatedLOs?type=sourceAlternateLOs&limit={n}

The response returns learning objects that can act as alternates and includes a meta.count field that indicates the total number of such alternates, independent of the current limit. Integrations can use this to present recommended equivalents, or to build administrative views of alternate mappings.

Reporting and analytics use cases

Users that generate reports or analytics should update their logic to add isAlternateComplete and alternateCompletions into their reporting workflows.

Reporting integrations that consume completion data (for example, LT export, custom data warehouse feeds, or BI dashboards) should extend their logic to read and persist both attributes.isAlternateComplete and relationships.alternateCompletions from the LO APIs:

Typical use cases:

Without incorporating these two fields, downstream reports will treat alternate completions as regular completions and will  not be able to explain   why  a learner shows as completed a training they never took directly.

Learner vs Admin LO API behavior

The multi language job aid structure is identical in both learner and admin LO APIs. Learner scope returns only those job aids visible to the learner, but for each visible job aid it exposes all configured locales via multiple resource entities (one per locale) and multi locale localizedMetadata. Admin scope returns all job aids the admin can manage, with the same LO model and locale specific resource IDs. Clients with learner scope should choose the resource whose attributes.locale best matches the learner’s content language, while admin tools can enumerate all locales for reporting and management.

Checklist with commenting capability

To support workflows where reviewers can share structured feedback on checklist‑based activities, this release surfaces  checklist comments  and reviewer visibility controls through the learning object resource API.

Checklist‑related metadata is exposed on learningObjectResource entities (JApiLOResource, “type”: “learningObjectResource”) that represent checklist resources within a course or other learning object.

The information is available via:

GET /primeapi/v2/learningObjects/{loId}?include=instances.loResources

When the learning object instance contains checklist‑type resources, the corresponding learningObjectResource entries in the included array expose comment and reviewer‑visibility attributes under attributes, and reviewer identity under relationships.

New checklist comment attributes

For checklist resources, the following attributes may be present on the learningObjectResource:

Other checklist‑specific context such as checklistEvaluationStatus, isChecklistMandatory, resourceSubType: “CHECKLIST”, and submissionDate are also available on the same learningObjectResource to support richer checklist UIs and reporting.

Reviewer identity relationship

When reviewer names are meant to be visible, the learningObjectResource includes a relationship that points to the reviewer user:

"relationships": {
  "checklistReviewedBy": {
    "data": {
      "id": "user_id",
      "type": "user"
    }
  }
}

This relationship is populated only if:

Client applications should:

Accessing checklist resources and comments

To retrieve checklist resources and their comments for a given learning object, clients should:

GET /primeapi/v2/learningObjects/{loId}?include=instances.loResources

This pattern allows headless or custom clients to render a comprehensive checklist experience, including status, mandatory/optional flags, and reviewer feedback, directly from the Prime APIs.

Reporting and UX considerations

Multi‑language support for job aid

To support implementations where job aids must be delivered in multiple languages to both learners and admins, this release extends job aid handling to return  one resource per locale  instead of a single hard‑coded en-US resource.

Multi‑language job aids continue to use the existing hierarchy:

learning Object (lo) →  learningObjectResource  (loResource) → resource

No changes are required to the API contract. Any localized job aid fits into this structure naturally, with separate resource entities per locale and shared localized metadata at the learningObject / learningObjectResource levels.

Job aid data is exposed via:

GET /primeapi/v2/learningObjects/jobAid:{jobAidId}?include=instances.loResources.resources

When a job aid has multiple language variants, the included array contains multiple “type”: “resource” entries—one for each locale (for example, en-US, fr-FR, es-ES), all linked from a single learningObjectResource.

Multi‑language job aid structure

Multi‑language job aids use:

For a multi‑language job aid, a typical pattern is:

Clients can select the appropriate resource by matching the learner’s locale to the resource.attributes.locale field.

New resource ID format for job aids

To properly distinguish multiple localized variants of a job aid resource, the resource ID format has changed.

Old (legacy) resource ID format

Previously, job aid resources used an opaque ID format such as:

jobAid:131032_-1_-1_2_resource

This format did not encode locale, and APIs would effectively expose only a single resource (usually en-US).

New resource ID format (multi‑language aware)

The new format is:

jobAid:<jobAidId>_<version>_<localeCode>

Examples:

Visual breakdown:

jobAid:131032_2_fr_FR

        │      │ │ │

        │      │ │ └──► Locale Code (fr_FR, en_US, es_ES, etc.)

        │      │ └────► Version (2)

        │      └──────► JobAid ID (131032)

        └─────────────► Resource Type Prefix ("jobAid:")

This structure lets clients:

Backward compatibility:

/resources/{resourceId}

The legacy resource endpoint remains available:

It is now  backward compatible  with both the old and new ID formats:

Integrations that currently store or reference the old resource IDs can continue to function without change, while newer implementations are encouraged to adopt the new ID format for locale‑specific operations.

Integration and UX considerations

Time‑slot constraints for starting modules

Some learning experiences must be available only within a defined time window. The release displays time‑slot information on learning object resources and introduces a validation endpoint that checks whether a learner can start a resource at the current time.

Time‑slot metadata is available via the endpoint:

GET /primeapi/v2/learningObjects/{loId}?include=instances.loResources

At the learning object resource level, a timeSlot object may now be present in the attributes, with startTime and endTime values in UTC. This specifies the window during which the resource may be started.

Before launching a module, integrations can call a new validation endpoint:

GET /primeapi/v2/learningObjects/{loId}/instances/{loInstanceId}/loResources/{loResourceId}/canStart

This endpoint, intended for learner‑read scenarios, returns whether the learner is currently allowed to start the resource, considering the configured time slot, delivery type and other back‑end rules.

Custom players and client applications should use canStart to enforce time‑based access and use the timeSlot metadata to display clear messaging about when content becomes available or expires. Negative responses from canStart should be handled explicitly to inform learners why the content cannot be started yet or anymore.

Multi‑attempt, content‑driven quizzes

Some content packages implement their own attempt tracking rather than relying solely on Adobe Learning Manager. The release introduces a flag that indicates when attempts are driven by the content itself.

Through:

GET /primeapi/v2/learningObjects/{loId}?include=instances.loResources

learning object resources may now expose a Boolean attribute hasContentDrivenAttemptTracking. When this is true, the quiz or module manages attempts internally (for example, via SCORM or xAPI logic), and the platform’s standard attempt counters may not fully reflect the learner’s experience.

Integrations that display attempt counts or control retry behavior should check this flag. When it’s enabled, they should not infer attempt limits solely from platform metadata and should be prepared to rely on content‑side reporting (for example, via xAPI statements) or business‑specific rules.

Behavioral change in Job Aid resource ID format

This release introduces an important behavioral change in the format of Job Aid resource IDs. While no new endpoint is involved, this has a direct impact on systems that store or parse these IDs.

Previously, Job Aid resource IDs used a format like:

jobAid:<jobAidId>_-1_-1_2_resource

In the April 2026 release, this is replaced with a simplified and more explicit format:

jobAid:<jobAidId>_<version>_<localeCode>

For example:

jobAid:131032_2_fr_FR

The components are:

Any integration that indexes resources or persists in Job Aid resource IDs must update its parsing and storage logic to recognize the new format. Because the identifiers themselves change, it’s strongly recommended that you rebuild any local indexes keyed by Job Aid resource IDs after upgrading to the April 2026 release.

Set course banner images via migration

You can now set or update course banner images in Adobe Learning Manager as part of your standard migration workflow. This helps you launch or clean up large catalogs while keeping your course pages visually consistent, without manually editing

Where banner images are defined

Banner images are configured in the course.csv migration file, which you already use to provide course metadata such as:

With this enhancement, the course.csv specification includes an additional optional column for the course banner image. The exact header name is defined in the CSV specification you download from your Learning Manager account (for example, it may appear as bannerUrl or bannerImageUrl).

The banner column:

Preparing banner images for migration

id,courseName,courseCreationDate,state,author,thumbnailUrl,bannerUrl
12345,DEMO1,2018-05-05T08:56:21.000Z,Published,Sudheer,pic1.png,banners/banner1.png
45678,DEMO2,2018-05-05T08:56:21.000Z,Published,Sudheer,pic2.png,banners/banner2.png

Apply banners during a migration run

Once your course.csv is updated, the flow is the same as any other migration:

New courses vs. existing courses

The banner field works in both scenarios:

Your actual column names and paths must match the downloaded CSV specification and your content repository layout.

Remove order column in LearningProgramCourse

Adobe Learning Manager now supports a simplified model for course ordering inside Learning Paths (Learning Programs) during migration. As part of this change, the order column is removed from the learning_program_course.csv migration file.

Previously, the learning_program_course.csv file included a column (often called order or similar) that was used to:

Now, Adobe Learning Manager no longer uses the order column in learning_program_course.csv to determine course ordering. Instead, the migration CSV focuses on which courses belong to which Learning Program, rather than how they are ordered numerically.

Impact on migration CSVs

learning_program_course.csv

You should treat the legacy order column as removed or ignored:

Always refer to the latest CSV specifications from your Learning Manager account (via csv_specifications.zip) to confirm the current header set and requirements.

timeZoneCode on Course Instances

From this release onwards, the Course Instance model (learningObjectInstance) exposes a new attribute:

timeZoneCode – a string field that explicitly links a course instance to one of the account’s configured time zones.

This allows clients to:

Where timeZoneCode appears

The field is added in the attributes of the learningObjectInstance model
for course instances only.

Example:

{
  "id": "course:1262748_1359228",
  "type": "learningObjectInstance",
  "attributes": {
    "dateCreated": "2019-08-06T13:50:39.000Z",
    "isAET": false,
    "isDefault": true,
    "timeZoneCode": "356",
    "isFlexible": false,
    "state": "Active",
    "localizedMetadata": [
      {
        "locale": "en-US",
        "name": "Default instance"
      }
    ]
  }
}

How to resolve timeZoneCode

The numeric timeZoneCode is a lookup key into the account’s time zone catalog, which is exposed via the Account API:

GET /primeapi/v2/account
Authorization: Bearer <access_token>

Within the response, time zones are listed in:

"data": {
  "attributes": {
    "timeZones": [
      {
        "name": "Etc/GMT+12",
        "timeZoneCode": "356",
        "utcOffset": -720,
        "utcOffsetCode": "UTC-12:00(Etc/GMT+12)",
        "zoneId": "Etc/GMT+12"
      },
      "..."
    ]
  }
}

Recommended client flow:

Asynchronous public APIs for user group membership

Introduction

Adobe Learning Manager exposes two admin async APIs to manage user group (UG) membership:

Instead of updating membership in the same request, these APIs accept your batch and return an event ID. The actual result is delivered later via webhooks.

Use them when:

For small, interactive admin actions, you can continue using the synchronous /userGroups/{id}/users APIs.

Authentication and base URL

These APIs are part of the standard v2 Admin API surface.

For general Admin API behavior and scopes, see:

Request format

Both add and remove use exactly the same body shape.

Body

{
  "metadata": {
    "event_id": "my-optional-correlation-id",
    "sourceSystem": "HRIS",
    "batchId": "hr_2026_03_30_0001"
  },
  "data": [
    { "type": "user", "id": "11101219" },
    { "type": "user", "id": "11101220" }
  ]
}

data (required)

data is the list of user resource identifiers for this batch.

Constraints

If any of these conditions fail, the API returns an error and nothing is queued.

metadata (optional)

metadata is any additional correlation data you want echoed back in the Webhook.

Typical usage:

The service returns this object unchanged in the Webhook response, so you can match the callback to your internal job.

Constraints:

Response format

Both endpoints respond synchronously with an event ID:

{ "event_id": "cd2972c8-cb15-47a0-a23f-e4a16cb720f5" }

Behavior:

You should always:

View Webhooks for adding and removing user group membership for more formation.

Frequently asked questions

How does the April 2026 release change the learningObjects API for adaptive learning programs?

This release adds an isAdaptive flag on learning programs and makes sections and subLOs learner‑aware. In adaptive programs, learners only see the sections and sub‑learning objects that are visible to them based on adaptive rules, while admins can see the underlying adaptive configuration for a user group.

Do I need to update my integration if it assumes all sections and sub LOs are always returned?

Yes. For adaptive learning programs, the API now returns only the sections and sub LOs that are visible to the current learner. Client code should treat the response as the authoritative, possibly filtered view, instead of assuming a complete list.

How can I programmatically reset a learner’s completion for a course or learning program?

Use the new endpoint:

POST /primeapi/v2/learningObjects/{loId}/instances/{loInstanceId}/refreshCompletion
This resets completion for the targeted instance when permissions and state allow it.

How do I tell if a learner completed something via an alternate or equivalent learning object?

Check the isAlternateComplete attribute on the learning object. If it’s true, the alternateCompletions relationship lists the learning objects that acted as alternates for that learner’s completion.

How can I find all alternates that can satisfy a particular learning object?

Use the following endpoint:

GET /primeapi/v2/learningObjects/{loId}/relatedLOs?type=sourceAlternateLOs&limit={n}

and use the data array (for the alternates) and meta.count (for the total number of alternates).

How do I know whether a learner is allowed to start a module at this moment?

First, fetch the resource’s timeSlot from:

GET /primeapi/v2/learningObjects/{loId}?include=instances.loResources
and then use:

GET /primeapi/v2/learningObjects/{loId}/instances/{loInstanceId}/loResources/{loResourceId}/canStart

What does hasContentDrivenAttemptTracking mean on a resource?

It indicates that attempt tracking is controlled by the content itself (for example, via SCORM/xAPI logic) rather than just by platform counters. When it’s true, don’t rely solely on platform attempt counts or limits for your UX and reporting.

How do I get menus appropriate for non‑logged‑in users (public experiences)?

Use:

GET /primeapi/v2/templates/menus?include=pages,subMenus.pages&isNonLoggedIn=true

This returns menu and page structures filtered for anonymous users, suitable for Experience Builder or other headless sites.

What has changed in Job Aid filtering with effectiveModifiedDate?

Requests that combine filter.effectiveModifiedDate with filter.loTypes=jobAid now correctly return only Job Aids within the specified date window.

What has changed in the Job Aid resource ID format and how should I handle it?

The ID format changed from values like:

jobAid:<jobAidId>_-1_-1_2_resource

to:

jobAid:<jobAidId>_<version>_<localeCode>

for example jobAid:131032_2_fr_FR. Any system that stores or parses Job Aid resource IDs must be updated, and you should plan to rebuild local indexes keyed by these IDs after upgrading to the April 2026 release.