What It Means to Build Trust in the Age of AI

Not-so-fun fact: it can take nearly 300 days to identify and contain a data breach.?

AI has changed how we approach our processes, operations, and productivity. It has also transformed the state of data security. Data breaches, phishing attacks, system vulnerabilities, and other threats to your data have been magnified with the emergence and evolution of AI. Consumer trust is at its lowest point in recent years, and organizations are catching on as 86% of IT leaders expect to increase their security investments.?

As we enter the new wave of agentic AI, data security is no longer a “nice-to-have” but a means of survival. Agentforce is a comprehensive agentic platform with everything you need to scale autonomous AI agents securely throughout your enterprise with low-code solutions. And it was designed with customer trust front and center.?

Salesforce’s Office of Ethical and Humane Use addresses trust concerns by codesigning and codeveloping ethical controls and processes across teams — from product and engineering, to Legal, Privacy, and more. The Office of Ethical and Humane Use is a fundamental part of the development process at Salesforce, ensuring that trustworthiness is at the core of everything, including Agentforce.

To help you build your own trustworthy Agentforce, here are our top takeaways.?

1. Reduce hallucinations with topic classification. A “topic” is a category of actions related to a particular job to be done by Agentforce. Topics contain “actions,” the tools for the job, and “instructions,” which tell Agentforce how to make decisions. Topics overall define what Agentforce can handle, and manage and redirect unwanted inputs and outputs. In turn, Agentforce doesn’t try to answer a question it shouldn’t, reducing the propensity to hallucinate.
2. Limit the frequency of agent-generated emails. Administrators can cap the frequency of AI outreach to provide better experiences for the recipient. Capping email frequency during setup prevents overreach, which reduces the likelihood of email fatigue and opt-outs.
3. Respect user privacy with opt-out features. Agentforce lets customers and prospects opt out of communications. This feature, integrated directly into our customer relationship management (CRM) software, provides a seamless customer experience, allowing customers to control how many communications and emails they receive.
4. Create transparency by design. To build trust, it’s important that users know when they’re interacting with an AI agent. By default, Agentforce products use standard language to alert administrators and agent managers when they’re about to implement or use an AI agent. These notes highlight the capabilities and limitations of AI, ensuring a clear understanding of its impact and potential.?
5. Ensure smooth AI-to-human handoffs. A successful implementation of agentic AI requires seamless transitions between AI agents and the human workers they support. For example, an Agentforce SDR can copy a sales manager on each of the agent’s email. This approach fosters an AI-human partnership in every interaction to build and maintain trust.

These trust patterns are implemented in addition to the Einstein Trust Layer, a foundational element of Salesforce’s AI systems. Within the Trust Layer is the Audit Trail feature, which lets users see what Agentforcedid, why it did it, and the outcomes of its actions.

In the age of AI, it’s either take a proactive approach to your security, or face costly fees and losing customer relationships. While Agentforce’s built-in features give you guardrails that protect the privacy and security of your data, and improve the accuracy of your AI results – each enterprise has its own unique security needs that need to be addressed. Whether your organization is looking to protect itself from data loss and corruption,? needs a real-time pulse on who has access to what data, or any other security challenges your organization faces — Salesforce’s trusted services portfolio has a solution for your business needs. By investing in security, learning security best practices, and prioritizing trust in all parts of your organization, you can be sure your organization is ready for AI.

Additional reading

Introducing: Agentforce Testing Center. Become an agent-first enterprise by quickly and efficiently testing, deploying, and monitoring your Agentforce at scale with this new suite of Agentforce Lifecycle Management tools.

4 Key Practices to Protect Your Data. A secure AI strategy is key to deepen customer relationships, drive innovation, and make smarter decisions.

How to Scale Test for Seamless Performance. Before you deploy — stress-test your Agentforce and AI apps to ensure smooth performance, even during peak traffic.

Security and Privacy for Data Cloud. Protect your Data Cloud data by identifying, encrypting, and anonymizing sensitive data with these Salesforce security solutions.

Top AI Agents Predictions & Trends in 2025.As we look toward the era of agentic AI, there are a many ways it may come to life — and strategies to ensure it will be successful. Salesforce leaders weigh in.

Upcoming events

• January 23, 2025: Improve Data Governance with Salesforce
• January 29, 2025: Encryption Made Easier with Salesforce
• March 5-6, 2025: TDX 2025