Ethical AI in Salesforce: Building Trust in the Age of Automation

In an era where artificial intelligence is redefining business processes, maintaining ethical standards is crucial. Salesforce, a pioneer in CRM solutions, is at the forefront of incorporating ethical AI practices to build trust and transparency. This article explores the ethical considerations surrounding AI in the Salesforce ecosystem and offers insights into fostering responsible automation.

Back in 2018, Salesforce embarked on a year-long journey to establish its trusted AI principles. This wasn't just a top-down approach; it was a collaborative effort involving contributors from all corners of the company – from engineering and product development to legal, equality, government affairs, and marketing. The result? A set of principles approved by executives across various roles and clouds, including co-CEOs Marc Benioff and Keith Block.

Salesforce is guided by five core principles:- trust, customer success, innovation, equality, and sustainability. Salesforce is also deeply committed to the responsible development and deployment of our technology, driven by our Office of Ethical and Humane Use.

Responsible AI: Asking the Right Questions

Salesforce begins its AI journey by posing critical questions: not just if something can be done, but whether it should. This responsible approach involves collaboration with external human rights experts to stay informed, learn, and continuously discover ways to protect human rights.

Salesforce aims to empower users by providing tools and resources to identify and mitigate biases in AI systems. This includes features such as flagging when protected data categories are used in a model and offering transparency into factors influencing predictions. The goal is to ensure that AI is adopted in a safe and reliable manner.

Security and privacy are paramount. Salesforce adheres to the highest standards, complying with applicable laws and striving for excellence in research, safety, and quality. Their commitment extends to sharing research through peer-reviewed publications, conferences, and industry events.

Accountable Engagement: Collaborative Learning and Improvement

Salesforce engages with external human rights and technology ethics experts through the Ethical Use Advisory Council. They actively seek feedback from customers through advisory boards and open dialogue, incorporating this input into their decision-making process. Collaboration with industry groups, civil society forums, and governmental organizations demonstrates their commitment to continuous improvement.

Employees are encouraged to voice questions and concerns through channels like anonymous online reporting, Slack channels, and group email addresses.

Transparent Practices: Illuminating the AI Black Box

Transparency is not just about building models but also explaining why they make specific predictions or recommendations. Salesforce publishes model cards detailing creation, use cases, ethical implications, and performance scores. Model explainability is provided when predictions or recommendations are made.

Customers remain in control of their data and models at all times. Salesforce ensures clear disclosure of terms of use and the intended applications of AI capability.

Empowering Accessibility: Making AI User-Friendly

Salesforce strives to make AI accessible to individuals of all technical skill levels. They simplify the complexity of AI, allowing not only advanced data scientists but anyone to build AI applications with clicks, not code. Free AI education via Trailhead equips users with the skills needed for the Fourth Industrial Revolution.

The Salesforce AI team is committed to delivering breakthroughs to inform new product categories, keeping customers at the forefront of technological advancements.

Inclusive Approach: Building Diversity into AI

Salesforce tests its models using diverse and representative data sets, aiming to understand the impact on a broad range of customers, end users, and contexts. Inclusive teams, aligned with the core value of equality, are built to represent diverse experiences and points of view.

Driving Transparency in Generative AI: A Commitment to Clear Communication

Salesforce prioritizes transparency throughout the development and deployment of its generative AI systems. This includes providing disclaimers, such as popups, to set expectations about the evolving nature of generative AI technology. Documentation, including model cards for predictive AI, offers customers key information and insights into AI models, training processes, and potential risks.

Responsible usage guidance ensures transparency in AI interactions, notifying users when biased inputs are detected, preserving ethical standards in the deployment of generative AI.

Empowering Users with Generative AI: Augmentation for Productivity

Salesforce's generative AI is a tool designed to support customers with augmentation and empowerment, helping individuals elevate their productivity and skill sets. For example, in Service Cloud, Salesforce's new generative AI Einstein Service Replies product proved highly effective in rapidly resolving issues, enabling service reps to explore new opportunities, such as upselling and cross-selling.

Salesforce ensures its AI technology remains user-friendly and inclusive, offering no-code and low-code products that democratize access to AI for people with varying levels of expertise. Prompt Studio, for instance, enables users to explore prompt engineering and create the best results while providing guidance on parameters and prompts.

Sustainable Generative AI: Reducing Carbon Footprints in Development

Addressing the environmental impact of generative AI development is a top priority for Salesforce. The conventional process of training large language models (LLMs) can result in a substantial carbon footprint, equivalent to the annual power consumption of over 1,000 U.S. households. To combat this, Salesforce is committed to sustainable practices, optimizing model sizes, minimizing unnecessary compute usage, and maximizing hardware efficiency—all while embracing renewable energy sources.

Conclusion: United in Ethical AI Advancements

At Salesforce, ethical AI isn't just a guideline – it's a commitment embraced by the entire Salesforce community. By navigating the complexities of AI with transparency, accountability, and inclusivity, Salesforce is not only shaping the future of CRM but also driving positive change across industries. In a collaborative spirit, Salesforce invites employees, customers, partners, and communities to contribute to the ongoing dialogue, ensuring that ethical considerations remain at the forefront of AI advancements.