Dreamforce opens Tuesday with AI pricing and governance in focus for enterprise buyers

Salesforce’s Dreamforce conference opens Tuesday in San Francisco, and the agenda is set to be dominated by artificial intelligence product and policy updates that affect how large companies buy and govern AI. The company is expected to outline roadmap changes for Einstein and Data Cloud, along with model access options from partners such as OpenAI, Anthropic, Cohere, and AWS, a continuation of strategies it highlighted in 2023 and 2024.

The timing matters because buyers are entering budget finalization for 2027 while new compliance obligations take hold. General purpose AI transparency and documentation requirements in the EU AI Act began phasing in about a year after the law’s 2024 entry into force, and procurement teams are now testing these rules against vendor offerings according to European Commission guidance.

What to watch on pricing and model choices Salesforce introduced Einstein Copilot and the Einstein 1 platform in 2023 and expanded access in 2024, positioning its AI as native to CRM and Service workflows. Customers will look for clarity on per user pricing, consumption-based add ons, and any credits for model inference that tie to Data Cloud usage, as peers such as Microsoft priced Copilot for Microsoft 365 at 30 dollars per user per month when it reached general availability in November 2023.

Model flexibility will be another theme. Salesforce has promoted a bring your own model approach through partnerships with OpenAI, Anthropic, Cohere, and Databricks. Amazon announced a commitment of up to 4 billion dollars to Anthropic in 2023, and Google disclosed a 2 billion dollar investment that same year, investments that keep the model supplier landscape in flux for enterprise platforms.

Compliance, trust, and where the data lives CISOs want proof of data residency, audit trails, and red teaming methods that satisfy EU AI Act and sectoral rules. The European Commission’s AI Office has begun standing up oversight processes, and companies are mapping controls to NIST’s AI Risk Management Framework published in 2023, along with ISO standards work such as ISO 42001.

Cost control will remain front of mind as AI scale tests meet reality. IDC has forecast worldwide spending on generative AI solutions to reach 143 billion dollars in 2027, up from single digit billions in 2023, which puts pressure on vendors to document measurable productivity gains inside sales, service, and marketing workflows.

For the broader tech landscape, the week is a barometer of how quickly AI will move from experiments to governed, line of business deployments inside core enterprise systems. If Salesforce pairs credible governance with clear pricing and flexible model access, it will signal a shift toward platform anchored AI buys that could shape cloud workloads across AWS with 31 percent market share, Microsoft Azure with 25 percent, and Google Cloud with 11 percent according to Synergy Research’s 2024 data.