Style Books in Liferay are now explicitly tied to a specific theme at the moment of creation, using the frontend token definition provided by that theme (via OSGi or themeCSS client extension). This structural link now ensures that each Style Book can only operate within the boundaries of its associated theme, eliminating cross-theme token contamination and enforcing clearer theme-based design governance.

Key Benefits:

• Users can no longer save Style Books that accidentally combine tokens from different themes, avoiding visual inconsistencies and design regressions.

• Every Style Book now visibly shows which theme it belongs to, reducing errors and making it easier for teams to manage design assets across multiple sites.

• When applying a Style Book to a page, the system will only list those created with the same theme as the page’s current one—no more trial-and-error or guesswork.

• If a Style Book becomes incompatible with the applied theme (e.g., after a theme change), it will be automatically unlinked to prevent display issues.

• During platform upgrades, existing Style Books are automatically linked to the site’s current public theme (as defined in Site Builder > Pages > Options > Configuration), reducing manual cleanup work.

• If a Style Book is imported without a valid themeId, the user gets a clear warning and knows exactly what’s missing to fix the import.

• Style Books based on themes that are no longer deployed or no longer provide a valid frontend token definition are automatically marked as inactive—clearly flagged and non-selectable.

• The OSGi or themeCSS client extension ID is displayed for inactive Style Books, helping devs or admins identify which theme needs to be re-installed or fixed.

With the release of Liferay DXP 2025.Q2, we are deprecating the Elasticsearch 7 compatibility due to the end-of-maintenance and upcoming EOL of the Elastic Stack 7.17. Liferay strongly recommends all customers with 7.17.x or earlier deployments to upgrade to the latest compatible version of Elasticsearch 8.x. Learn more. 

The Concurrent and Sync Reindex Modes, providing high-availability, introduced in Liferay DXP 7.4 Update 98 / Liferay DXP 2023.Q4 are now GA.

Reduce downtime, improve operational and resource utilization efficiency with the new Concurrent and Sync reindex execution modes providing high-availability in search capabilities while the operation is running.

Managing and synchronizing content across multiple environments just became easier and more reliable. With this release, Liferay expands the capabilities of the Batch Engine by introducing two powerful enhancements:

1. Batch Delete by External Reference Code – Users can now delete items using external reference codes instead of internal IDs, making batch deletions simpler and more consistent across environments.

2. Import Strategy Support for Deletion – When performing a batch delete, users can now define whether the operation should be partial (stop on error) or complete (continue even if some items fail), offering more control and fault tolerance during the process.

Together, these updates streamline how teams manage bulk deletions—especially in environments that rely on external identifiers or require consistent data maintenance across staging, production, and other instances.

Key Benefits:

• More reliable environment synchronization: External Reference Codes allow you to delete the same entities across different environments without depending on internal IDs, reducing risk of mismatches.

• Simplified bulk deletion workflows: Deleting large sets of data is now easier, with fewer manual steps and lower chance of errors.

• Greater control over delete operations: Choose whether the process should stop on errors or complete fully—helping teams tailor the behavior to fit their operational needs.

• Increased resilience and fault tolerance: Deletion jobs are less likely to fail entirely due to minor issues, ensuring smoother maintenance processes.

• Consistent support across entities: These enhancements are available for all entity types supported by the batch engine, making them broadly applicable across different use cases.

We made sure to be able to configure the most relevant SCIM providers by implementing their specifications(Microsoft Entra, Cyber Ark, etc.)

Key Benefits:

Implementing SCIM endpoints from top identity providers streamlines and secures user management by enabling automated provisioning, deprovisioning, and syncing of user data across platforms. This enhances security, ensures compliance, reduces manual work, and improves consistency and user experience. With standardized interoperability, it also future-proofs integrations and boosts operational efficiency while offering better visibility and control over identity lifecycle events

In order to add more flexibility to every execution in batch engine, we have added a new parameter ( batchExternalReferenceCode ) to be capable to send it to the backend without updating the ERC of the import task.

Key Benefits:

• Simplicity of execution with one parameter for each purpose.

• Add new capabilities for all entities that supports batch.

Captcha engine needs to be configurable on instance level to satisfy the multi-tenant environments' need. That has to be possible to configure them for one instance without interfering with the configuration for other instances.

Key Benefits:

Enabling CAPTCHA at the Instance level adds flexibility in configurations without affecting global settings. This supports tailored security and user experiences, and enables compliance per instance, all while retaining centralized control when needed

It’s going to be available without enabling the BETA flag, and from now it’s officially supported.