Highlights
- AI tools like Setup with Agentforce and Agentforce Vibes are changing the way that Salesforce professionals build on the platform. However, they don’t replace the need for human orchestration.
- Salesforce itself is doubling down on Trailhead by adding Agentblazer Status and more Agentforce content to its existing Certifications.
- AI tools are not perfect – they should be used to build a starting point, but they don’t replace the need for a human to verify and test thoroughly. A human will also own the output.
- Fundamental Salesforce knowledge is arguably more important than ever in the AI era. AI tooling can build beyond your understanding of the platform, so it’s critical to make sure you know more of Salesforce’s inner workings.
For over a decade, Salesforce Trailhead has been the go-to platform for learning Salesforce. From total beginners to seasoned Salesforce veterans, Trailhead has something for everyone. Whether you’re trying to learn what a CRM system is, or you want to see how JavaScript works in a Lightning Web Component, Trailhead is your go-to resource.
However, the question has been raised: now that various AI tools are available that can do the work directly in Salesforce on your behalf, is Trailhead still necessary? Do I really have to learn how to do things in Salesforce, or can I just use AI to do it for me? This article will explore the current state of Trailhead, as well as whether or not it is still a necessity for Salesforce professionals.
Salesforce Is Doubling Down on Trailhead
Despite pushing its new agentic capabilities over the last few years, Salesforce has been doubling down on educating humans using Trailhead. This shows that the company knows that humans will still be a necessary part of the Salesforce ecosystem. Fancy AI tools can make things easier, but they can’t replace the human that knows Salesforce properly.
In fact, Salesforce has made it clear that it wants those building on the platform to know Agentforce intimately. In February 2025, shortly after the original launch of Agentforce, Salesforce released Agentblazer Status on Trailhead as a way of helping people learn how to work with Agentforce.
Not long after that, Salesforce also updated the weightings of its Salesforce Admin Certification to include new Agentforce categories. This meant that Salesforce’s expectation for its future Salesforce Admins is that they’ll do more with Agentforce, and less configuration of the Salesforce platform directly themselves.
At the end of 2024, Salesforce announced that it was going to be making two of its AI Certifications free for everyone, and provided training resources on Trailhead to help people get up to speed.
Despite all of this evolution in the AI space, Salesforce still continued evolving its Trailhead platform to help people learn. Even when AI tooling was taking over the conversation, Salesforce knew that the best way to upskill its global community of builders was to continue to educate them through Trailhead.
What AI Can Do in Salesforce Today
There’s no denying that AI is doing more for Salesforce professionals than ever before. Lately, I’ll wake up in the morning, grab my phone, and immediately see some fancy new AI tool that could help me with my work in some way. New models and reasoning engines come faster than ever before, each making the way AI builds better and more practical than the day before.
In the Salesforce world, this is no different – in fact, AI is being adopted top-down by Salesforce itself. The recent AIforce hype at Dreamforce and the announcement of the Salesforce-Anthropic Claudeforce partnership are just the tip of the iceberg here.
Salesforce Admins have access to tools like Setup with Agentforce, which allows admins to communicate in natural language with their Salesforce orgs. It’s as simple as providing a prompt to Agentforce and watching it walk through the steps to complete the task.
Admins can also use AI to build flows, either through Agentforce or by using an LLM that leverages the Salesforce DX MCP. Troubleshooting is a breeze, too. AI tools can help admins discover what an Apex class does without having to reach out to a developer every time.
Speaking of Salesforce Developers, they’re reaping the benefits of AI as well. Tools like Agentforce Vibes, an AI-powered development experience built by Salesforce, can help developers build with ease. Whether it’s working with React, Lightning Web Components, Apex, or anything else, Agentforce Vibes can help by drafting code for you. You can even leverage Agentforce Vibes to build your agents for Agentforce. Agent-ception!
Then, there’s the Salesforce DX MCP, which empowers any supported LLM tool to build using over 60 different Salesforce developer tools. This means tools like Claude Code, Cursor, Codex, or Windsurf can help you build in Salesforce. If you’re not a fan of Agentforce Vibes, no worries! Pick your favorite agentic coding tool, connect it to the Salesforce DX MCP, and you’re off to the races.
There’s really no denying the impact that AI has had on the Salesforce builder ecosystem. I’ve not even spoken about Salesforce end users, who are benefiting from even more AI-powered tools. If you have been hesitant to adopt AI in your Salesforce work, now is the time to get on board. It’s clear that these tools aren’t going away anytime soon and are constantly evolving. Don’t miss out!
Just Because AI Can, That Doesn’t Mean It Should
One big caveat to the AI arms race that has a major impact on the human-agent relationship is that AI makes mistakes. Earlier LLM tools were known for frequently hallucinating, and while this has improved, it’s still a problem. Human beings must still validate that the output of an agent is accurate, functional, and achieves the desired outcome.
To do this properly, those human beings will need to have an understanding of the Salesforce platform that they and their AI tools are building on. AI tools are designed to create a starting point, not a finished product – this is Salesforce’s own explanation. This means that humans are still a critical and valued part of any Salesforce project.
Take Setup with Agentforce, for example. While it can take action on a request, it’s not going to proactively do anything in a Salesforce org without direct instruction from a human. It will even wait for a user confirmation after a solution to a prompt has been proposed before taking action on it.
Becoming an Orchestrator
The idea of changing Salesforce roles is nothing new. It’s something that we at SF Ben have talked about extensively, and we’re not the only ones doing so. The reality is this: AI is here, and it’s not going anywhere. Every Salesforce professional has had their role impacted in some way, shape, or form by the AI revolution.
Admins are expected to have their heads across new tools and know how to put them to use in their Salesforce orgs. Developers build faster than ever before, which is only enabled by the use of AI tooling for code writing, testing, and more. The reality is that these roles, along with every other Salesforce role, are changing. It means less on-the-tools work, and more of a focus on orchestration of AI tools.
What this means is that we all need to become better at our soft skills and our ability to communicate with both humans and machines. This means that, in addition to getting hands-on with the technology (albeit in a lighter way than before), we need to start:
- Clearly designing and communicating requirements and outcomes.
- Selecting the right tools for a particular job.
- Validating and verifying the output against the desired outcome, and how the Salesforce platform actually works (including making sure best practices are being followed properly).
- Testing the results of the outputs as well – not just checking if a flow is built the way we expect it to be, but making sure it does what it’s supposed to do, the way we expect it to be done.
- Ultimately, owning the outcome of any AI tool. Ownership must always belong to a human; there simply is no delegating ownership to an AI tool.
Survey data from across all our 2026 SF Ben surveys show that roles are changing, and AI is playing a bigger role. This is why it’s critical to learn how to work with AI properly. The answer is not to stop learning Salesforce and trust that AI tooling can do it for you, but instead continue to master the Salesforce platform and learn how to leverage the power of AI as an orchestrator.
Why Trailhead Is Still a Staple of Salesforce Training
It should be clear from the above that Trailhead, and more specifically mastery of the Salesforce platform itself, is still a requirement in the AI age. I’ve said it time and time again: if an AI tool can build something that you’re not fully across, you should be very careful deploying it. If something goes wrong, you’re responsible, and you must find the solution – AI tooling or not.
Today, your fundamental understanding of Salesforce is arguably more important than it was pre-AI. In the past, you’d be responsible for what you built. You’d only work with tools you knew well. Today, AI tools can build beyond what you may understand, and you’re still responsible for those outputs.
That’s not to say you should neglect your AI knowledge in favor of Salesforce mastery, but rather that you’ll need both, together. Learn the ways of Salesforce, and learn how to effectively put AI to work in that space.
Final Thoughts
Trailhead has been, and will be for the foreseeable future, a staple of Salesforce training for all Salesforce professionals. The era of AI building doesn’t do away with the need for human intelligence in the Salesforce ecosystem, but rather introduces new challenges that increase the need for it.
Your AI tools should draft the solution for you, but they should never make decisions for you. It should never create anything that you don’t fully understand, because you are still responsible for the outputs.
TLDR: Keep upskilling on Trailhead! It’s not going anywhere, and Salesforce’s content team is always pushing new and valuable content for you to upskill with. It’s a great place to learn Salesforce, but also a fantastic tool to learn how to use AI properly on the Salesforce platform.
