Release notes, new features, deprecations, and announcements from Salesforce and AWS.
68
Total
20
Salesforce
45
AWS
20 notices for Salesforce
Data Cloud as a CDP: DMOs and DLOs, four ingestion methods, identity resolution with match and reconciliation rules, segmentation, activation targets, and why Data Cloud is the foundation for Agentforce and all Salesforce AI.
Summer '26 sandbox preview is live. Key highlights: Agentforce for Developers (AI-powered code generation inside VS Code), enhanced Flow orchestration with sub-flows, Data Cloud real-time segmentation improvements, and the new Einstein Trust Score dashboard for monitoring AI quality across your org.
Flow Orchestration, previously a paid add-on, is now included as a standard Flow type for all eligible Salesforce editions. Customers can build multi-step, multi-user business processes with stages and steps without purchasing a separate license.
Apex now supports the null-coalescing operator (??). Write `String name = account.Name ?? 'Unknown';` instead of ternary checks. Also available: null-safe navigation combined with coalescing (`contact?.Account?.Name ?? 'N/A'`). Available in API version 66.0+.
Spring '26 brings Agentforce enhancements, expanded Einstein AI features, Dynamic Forms GA for all objects, and updated permission model with enhanced profile restrictions.
For new integrations, Salesforce now recommends using External Client Apps instead of Connected Apps. Connected Apps still work but External Client Apps provide better lifecycle management and security controls.
Spring '26 adds Flow Logging for native runtime monitoring and troubleshooting, AI-powered flow generation reaching GA so admins can describe a process in plain language and get a draft flow, inline editing and column sorting in Flow Data Tables, and orchestration debug improvements that let developers set custom start and end points to test specific segments of long orchestrations.
Workflow Rules and Process Builder reached end of support on December 31, 2025. They still run but receive no bug fixes or support. Migrate to Flow Builder using Setup > Migrate to Flow.
New certification validating expertise in building and deploying AI agents with Agentforce. Covers agent configuration, prompt templates, action libraries, and guardrails. $200 exam fee.
Winter '26 highlights include Agentforce for IT Service with native Slack integration, AI-powered decision routing in Flow Builder, multi-column list view sorting reaching GA, and an increase in Activity object custom fields from 100 to 300.
Summer '25 delivers hundreds of Agentforce enhancements including a new Agent API for triggering agents from external systems, Data Cloud Zero Copy Data Streams for live ingestion from Snowflake and BigQuery, and an Agentforce Testing Center that auto-generates test suites from natural language descriptions. The release also rebuilds the Send Email flow action with a rich-text WYSIWYG editor.
Hyperforce expanded to over 38 global regions in 2025, up from 4 at launch, with 90% of Salesforce customers now eligible for migration. New regions include additional Asia Pacific and European data centers, giving customers greater control over data residency and compliance.
Salesforce completed its phased MFA enforcement rollout, requiring multi-factor authentication for all users accessing Salesforce products via the UI. API-based and OAuth-based integrations are exempt. Orgs that had not enabled MFA voluntarily had it auto-enabled as part of the platform-wide enforcement.
Salesforce renamed Einstein Copilot to Agentforce as part of its broader shift toward autonomous AI agents, with the feature now appearing in Setup as Agentforce (Default). No functionality changed, but UI labels, permissions, and help documentation were updated to reflect the new branding introduced at Dreamforce 2024.
Salesforce launched Foundations, a free upgrade available to all Sales Cloud and Service Cloud Enterprise Edition and above customers, bundling Data Cloud, Marketing, Commerce, and cross-department features at no additional license cost. The Data Cloud allotment includes 10,000 segmentation and activation credits, automatic data harmonization for AI readiness, and a unified customer profile that feeds Agentforce. Email campaigns are included up to 2,000 sends per month. Foundations replaces the need to purchase separate Data Cloud licenses for customers who want basic data unification without a full Data Cloud contract.
As Salesforce migrated customer orgs to Hyperforce, Apex workloads benefit from elastic compute scaling, faster sandbox provisioning via Quick Clone, and zero-downtime major release upgrades introduced in Summer '24. Orgs on Hyperforce also gain access to Apex Guru and Scale Center, AI-powered tools that detect and surface code-level performance anti-patterns at runtime.
Dynamic Forms, which lets admins place individual fields and sections anywhere on a Lightning record page and control their visibility with filter conditions, reached general availability for standard objects in Winter '24. Previously limited to custom objects and a small pilot set of standard objects, Dynamic Forms expanded to hundreds of LWC-enabled standard objects including Account, Contact, Opportunity, Case, and Lead. Admins can now migrate existing page layouts to Dynamic Forms using the one-click migration wizard in Lightning App Builder, eliminating the need to maintain separate compact and full layouts and enabling field-level conditional visibility without code.
At Dreamforce 2023, Salesforce announced the Einstein 1 Platform, a rebrand and architectural consolidation of its core product suite around three pillars: Data Cloud for real-time customer data unification, Einstein AI for predictions and generative capabilities, and Flow for automation. Data Cloud was included at no additional cost for Enterprise Edition and above customers. Einstein Copilot Studio was introduced as a low-code builder for custom prompts, skills, and AI models. The Einstein 1 branding unified what had previously been marketed as separate products -- Tableau, MuleSoft, Slack, and the core CRM -- under a single platform narrative backed by the Salesforce metadata framework.
Salesforce made the Einstein Trust Layer generally available alongside Sales GPT and Service GPT, providing a set of built-in security guardrails for generative AI including dynamic grounding with CRM data, data masking before prompts leave Salesforce, zero data retention with LLM providers, toxicity detection, and an audit trail. The Trust Layer ensures that prompts and responses are never stored by third-party model providers or used to train external models.
Salesforce launched DevOps Center as a free, native change and release management tool that replaces change sets with a Git-backed, pipeline-driven workflow integrated directly into Setup. Teams can track work items, promote changes through environments via a visual pipeline, and connect to any Git provider, giving both click-based admins and pro-code developers a shared release process.
