“EGI Is Our True North Star”: What’s Coming Next for Salesforce AI?

Salesforce has just revealed the next stage of its venture in artificial intelligence, placing an emphasis on striving to create an EGI, or enterprise general intelligence. 

Alongside this, the SaaS giant has also provided more details on how it’s researching and applying “ambient intelligence” to Salesforce, which will shape how the next generation of agents work not only autonomously but proactively, in the background of any workflow. 

Salesforce Announces the AI Foundry 

As part of its AI research efforts, Salesforce announced its AI Foundry today – an extension of its AI Research arm that has and will continue to focus on three core research areas. It has been created to hone Salesforce’s research in the areas it thinks have the most development potential, and Silvio Savarese, Chief Scientist at Salesforce, has a notable role in both the Foundry’s inception and operation. 

“The problems that matter most for businesses don’t live at the model level anymore,” he said. “They live at the system level, where components work together to deliver accuracy, consistency, and reliability at scale. AI Foundry is the engine we’ve built to make that a reality.”

Which AI Areas Is Salesforce Betting On?

As aforementioned, the AI Foundry is focused on a select number of topics, specifically those that Salesforce believes are pertinent to the next steps of enterprise AI. 

Itai Asseo, the VP of AI Research at Salesforce, confirmed these areas are known as the company’s “big bets”. 

“These are the big bets that we’re focused on for this year in terms of taking this foundational research,” he explained. “We have simulation environments, agent-to-agent communication, and ambient intelligence.”

• Simulation Environments: In November, Salesforce launched eVerse, an enterprise simulation environment for voice and text agent training. Recently, eVerse was used to stress-test Agentforce Voice across thousands of simulated conversations, and Salesforce is now hyper-focused on developing agents that learn from experience.
• Ambient Intelligence: This refers to AI that is context-aware, proactive, and timely. It gains a complete understanding of the situation, preempts needs, and responds with little to no intervention, running in the background. Salesforce says its goal has always been to create “AI that’s always on but never overwhelming”, and this is a big part of that.
• Agent-to-Agent Communication: The AI Foundry is focused on developing an enterprise multi-agent semantic layer, which includes standardized protocols, guardrails, decision logging, and coordinated escalation. A key part of this investment is the creation of protocols, such as agent cards, to support this development.

Salesforce’s AI Research team has been active for a decade and has underpinned some of the company’s most significant AI releases and advancements, including Einstein GPT and Agentforce. Now that Agentforce has been generally available for over a year, the research focus has narrowed, deciding on what agents will look like in the near future and beyond.

Going forward, it’s clear to see that Salesforce is confident in the shift from generative to agentic AI, now moving further into autonomous and reactive AI, working a lot more with instinct based on context. 

All Roads Lead Back to EGI 

With AI research more important than ever before to the CRM giant, Salesforce has revealed it leads back to one main focus: EGI. 

“It’s very important that as we push the envelope of agentic AI,” said Silvio. “This is what we have been calling Enterprise General Intelligence, or EGI, which is our true north star for 2026-2027.”

EGI shares similarities with AGI – artificial general intelligence – but is more concentrated on AI systems that are “designed for business applications, and offer reliable and consistent performance across complex business scenarios.” At its core, it’s AI that can understand, reason, and act across an entire enterprise, not just in narrow use cases.

AGI, although often a hot topic of conversation for tech leaders, is still largely theoretical. EGI is already much more tangible. 

The potential impacts of Salesforce reaching EGI first are immense. Essentially, it could shift Salesforce from being a system of record to a system of action and decision-making. It would mean that Salesforce stops being just “the CRM” and becomes a solution that is able to action processes without being explicitly told to do so. This could be monumental for small/medium businesses (SMBs) or organizations that don’t have the staff to handle every task. 

Final Thoughts

If Salesforce can turn EGI from ambition into reality, it won’t just be advancing its AI strategy – it will be redefining its role in the enterprise stack, transforming Salesforce from more than just a CRM to an intelligent, reactive solution. 

Evidently, it needs to do a lot of work to get there, but the direction is clear: create the leading enterprise AI solution or risk being swallowed up by its competitors.