Combined AI Search release notes for upgrades from Washington DC to Xanadu

How to use this page

To help you prepare for your upgrade, we have combined the cross-family AI Search release notes onto one page. Read this summary of the new features, changes, and updated information for your product from Washington DC to Xanadu.

Tip:

If there were no updates for a release notes section in a certain family release, we included a short note for your reference. For example, if a product did not have any updates in Tokyo, the row says "No updates for this release."

Important information for upgrading AI Search to Xanadu

Before you upgrade to Xanadu, review these pre- and post-upgrade tasks and complete the tasks as needed.


Release	Release notes
Washington DC	Washington DC Patch 9: After you upgrade to Washington DC Patch 9 from an earlier release, make knowledge block content searchable by reindexing all your indexed sources that include knowledge articles. For details on reindexing, see Index or reindex an indexed source or Index or reindex multiple indexed sources. Washington DC: When you upgrade to Washington DC, AI Search automatically updates your existing Genius Result configurations to use the new AI Search Genius Result Configuration form fields. This update procedure makes the following changes: Removes existing Genius result answer type field values. Migrates Genius result logic field values to the new AI Search request processor and AI Search response processor fields as appropriate. After you upgrade your instance to Washington DC, AI Search retains the value that you previously set for the Boolean search operator to use when a search query includes multiple terms ( glide.ais.query.search_operator ) system property. To gain the benefits of the new enhanced query mode for multi-term searches, set this system property's value to AND then OR 2+ key terms. For details on AI Search system properties, see AI Search system properties. Starting in Washington DC, the User [sys_user] table defaults to sorting indexed records by their sys_created_on dates instead of sorting them by their sys_updated_on dates. This change requires reindexing of the User table indexed source for AI Search, which can be time-consuming. When you upgrade to Washington DC from a previous family release, AI Search does not automatically reindex the User table indexed source. If you need to be able to search the latest configuration for user records, you can manually reindex the User table indexed source, which may take some time. Otherwise, AI Search will reindex individual User table records as they're updated until all records have been reindexed.
Xanadu	Xanadu Patch 3: After you upgrade to Xanadu Patch 3 from an earlier release, make knowledge block content searchable by reindexing all your indexed sources that include knowledge articles. For details on reindexing, see Index or reindex an indexed source or Index or reindex multiple indexed sources. Xanadu: After you upgrade to Xanadu from an earlier release, perform the following steps to add the Dashboards, data visualizations, and KPIs navigation tabs to global search results in AI Search for Next Experience: Update the AI Search for Next Experience ServiceNow Store application to version 4 or later. For update instructions, see Update an application. Commit the update set provided in the AI Search for Next Experience 4.0 PAR tables update sets (KB1644544) article in the Now Support Knowledge Base. To learn more about update sets, see System update sets.

New features

Between your current release family and Xanadu, new features were introduced for AI Search.


Release	Release notes
Washington DC	(Washington DC Patch 9) Knowledge block content indexing When you index knowledge articles for search, AI Search now includes content from knowledge blocks to improve search recall. AND then OR 2+ key terms option for multi-term search queries Improve search performance for queries with two search terms using the new AND then OR 2+ key terms value for the Boolean search operator to use when a search query includes multiple terms ( glide.ais.query.search_operator ) system property. When this value is set, AI Search only performs automatic query resubmission if the search query includes more than two search terms after stop words are removed. AND then OR 2+ key terms is the default value for instances zBooted in Washington DC. Natural Language Query Genius Results Surface relevant search results from large source tables by adding Natural Language Query (NLQ) Genius Results to your AI Search applications. NLQ Genius Results interpret your searches using plain language to find source tables with relevant results. The NLQ Genius Result answer card displays the suggested source tables, combining results from Analytics Overview and CMDB table queries. You can preview records from each suggested table directly from the Genius Result answer card. Semantic vector search Find results that match the meaning and intent of your search terms. Semantic vector search is an alternate search mode that overrides the default lexical search mode. Now Assist in Virtual Agent uses semantic vector search to retrieve Catalog Items and topics. Now Assist in AI Search uses semantic vector search to find cached answers for Now Assist Q&A Genius Results.
Xanadu	(Xanadu Patch 9) Improve semantic search with third-party embedding models Use custom and third-party embedding models supported by the AI Search RAG application to generate more accurate and relevant semantic search results. (Xanadu Patch 3) Knowledge block content indexing When you index knowledge articles for search, AI Search now includes content from knowledge blocks to improve search recall. NLQ Genius Results support searching for multiple CMDB tables When your search includes terms that could match multiple CMDB tables, NLQ Genius Result answers display a drop-down list of potential CMDB table matches for each affected term. Use these drop-down lists to select the CMDB tables that you want to surface information for. Middle dot support for Japanese Increase search precision for Japanese phrases that include nakaguro (中黒, middle dot) characters. When indexing Japanese text or searching in Japanese, AI Search preserves individual terms separated by a nakaguro instead of merging them into a single term.

Changes

Between your current release family and Xanadu, some changes were made to existing AI Search features.


Release	Release notes
Washington DC	Synonym matching and expansion retains terms defined as stop words When evaluating synonym matches and expansions, AI Search retains search query terms defined as stop words. This behavior helps ensure that synonym matches return consistent results no matter which version of a synonym a user searches for. Field changes on AI Search Genius Result Configuration form On the AI Search Genius Result Configuration form, the Genius result answer type choice field has been removed. The Genius result logic script field has been replaced with separate AI Search request processor and AI Search response processor script fields.
Xanadu	Facets display refinement filters in the search user's language When displaying reference field values as facet refinement filters, AI Search uses field value translations for the search user's language if available. If no translations are available for the user's language, AI Search uses translations for the system language. Late binding security implementation preserves correct click rank search signals When late binding security removes inaccessible results, click rank search signals correctly reflect the final result ordering.

Removed

Between your current release family and Xanadu, some AI Search features or functionality were removed.


Release	Release notes
Washington DC	No updates for this release.
Xanadu	No updates for this release.

Deprecations

Between your current release family and Xanadu, some AI Search features or functionality were deprecated.


Release	Release notes
Washington DC	Starting with the Now Assist in AI Search 8.0 release, the External Content Q&A Genius Results feature is being prepared for future deprecation. It will continue to be supported until it is deprecated. For details, see the Deprecation Process [KB0867184] article in the Now Support Knowledge Base.
Xanadu	Starting with the Now Assist in AI Search 8.0 release, the External Content Q&A Genius Results feature is being prepared for future deprecation. It will continue to be supported until it is deprecated. For details, see the Deprecation Process [KB0867184] article in the Now Support Knowledge Base.

Activation information

Review information on how to activate AI Search.


Release	Release notes
Washington DC	AI Search is a ServiceNow AI Platform feature that is active by default.
Xanadu	AI Search is a ServiceNow AI Platform feature that is active by default.

Additional requirements

If any additional requirements were introduced or changed for AI Search we have noted them here.


Release	Release notes
Washington DC	No updates for this release.
Xanadu	No updates for this release.

Browser requirements

If any specific browser requirements were introduced or changed for AI Search we have noted them here.


Release	Release notes
Washington DC	AI Search does not support Internet Explorer.
Xanadu	AI Search doesn’t support Internet Explorer.

Accessibility information

Review details on accessibility information for AI Search, such as specific requirements or compliance levels.


Release	Release notes
Washington DC	No updates for this release.
Xanadu	No updates for this release.

Localization information

If there are specific localization considerations for AI Search we have noted them here.


Release	Release notes
Washington DC	AI Search supports international languages. For details of language support by feature, see Internationalization support.
Xanadu	AI Search supports international languages. For details of language support by feature, see Internationalization support.

Highlight information

If there are specific highlight considerations for AI Search we have noted them here.


Release	Release notes
Washington DC	Washington DC Patch 9: Expand search recall by indexing content from knowledge blocks. Washington DC: Display concise, actionable answers for plain language searches using Natural Language Query Genius Results. Increase search performance by skipping automatic query resubmission for searches with two search terms. Improve recall and consistency for synonym searches by retaining search terms defined as stop words. See AI Search for more information.
Xanadu	Xanadu Patch 3: Expand search recall by indexing content from knowledge blocks. Xanadu: Surface information across multiple CMDB tables with NLQ Genius Results. Increase search precision for Japanese phrases that include nakaguro (中黒, middle dot) characters. Scroll through multiple Genius Result answer cards in a carousel to select the most actionable answer for your question. See AI Search for more information.