Salesforce and AWS are widening their enterprise AI partnership so that customer data and AI agents show up inside the software people already use, rather than in separate portals. Salesforce context now surfaces in Amazon Quick without any custom integration work, and AWS agents are headed into Slack, beginning with the AWS DevOps Agent. The two companies announced the expanded collaboration this week, and neither disclosed pricing nor general availability dates for any of the integrations.
For Agentforce users, the headline change is model choice. Customers gain access to frontier models through Amazon Bedrock, with Anthropic and NVIDIA models available inside Salesforce agents and agentic workflows. OpenAI models are coming soon to Agentforce, and are already generally available on Bedrock. Agentforce Voice also interoperates with Amazon Connect Customer in real time, with Agent2Agent support enabling two-way live voice.
Read more: Salesforce Debuts AIforce to Push Its Data Into Any AI Interface
The data plumbing is the quieter part of the deal. Zero copy, Salesforce's approach to reading data where it lives, now reaches AWS Glue-managed Apache Iceberg tables and S3-backed Iceberg tables, and adds Amazon Aurora, Amazon RDS, and SageMaker Lakehouse. The point is cross-environment access without moving or duplicating data. Salesforce's headless architecture relies on the open Model Context Protocol standard, which lets the two products interoperate natively. Quick surfaces pipeline health, account summaries, service cases, and operational analytics.
Beyond the DevOps Agent, AWS plans to bring its Security Agent, FinOps Agent, and Partner Central agents into Slack, where they will have full conversational context. Salesforce says its Trust Layer adds dynamic grounding and toxicity detection, and the announcement cites HIPAA, PCI, SOC 2, and ISO 42001 compliance with full audit trails.
The announcement arrives with a roster of customer names attached. DXC has more than a dozen AI agents deployed and growing, according to the company. Executives from Engine, GoodRx, and Quicken Inc are also named in the announcement.
This is the latest in a run of Salesforce AI announcements. Earlier this month, Salesforce added seven job-ready agents to Agentforce, six of them generally available, with Hunter in pilot and scheduled for general availability in November 2026. That earlier batch also came with no pricing disclosed. Salesforce has said it delivered 7 billion Agentic Work Units across Agentforce and Slack, 3.2 billion of them in the second quarter alone.
AWS has been pushing its own agent products in parallel. In April, it turned internal supply chain and hiring tools into new AI agent offerings. Amazon Quick, meanwhile, picked up a desktop app plus Free and Plus plans, and already connects to Google Workspace, Microsoft 365, Zoom, and Salesforce.
What the announcement does not say matters too. There is no price for any of these integrations, no general availability date, and no regional availability detail. The partnership is described as an expansion, but readers looking to budget or schedule a rollout will have to wait for numbers.
