In this BSA series – “AI@Work” – enterprise software leaders explain in their own words how artificial intelligence (AI) is having a positive impact on people’s jobs and the workforce. In this submission, HubSpot’s Chief Legal Officer Alyssa Harvey Dawson writes about how HubSpot uses its AI ethics principles to ensure that a proper framework is in place to guide AI product development with a focus on effectively developing and implementing enterprise solutions.
At HubSpot, we’re proud to be recognized as a top workplace globally. We’re focused on building two products: one for our customers, and one for our employees, which is our culture (in fact, last quarter we published version 38 of our culture code). We are constantly iterating to provide the best experience for our employees and our customers – and that includes democratizing AI@Work.
HubSpot runs on HubSpot, and we are a testing ground for AI. We are actively experimenting, iterating, and innovating with AI to improve efficiency of HubSpot’s marketing, sales, and service. When we learn what works, we share best practices with our customers so they can scale.
While we recognize the value of AI to help businesses grow, we are equally mindful of AI risks. As risk mitigation measures, we have established policies and guardrails to guide responsible use and development of AI. Vendor contracts incorporate governance clauses to protect customer data from training AI models, and we conduct AI trust reviews to minimize privacy, security, and legal risks.
What distinguishes AI@Work from other uses of AI?
At HubSpot, our mission is to help millions of organizations grow better by democratizing access to sophisticated technologies, including AI. Over the past 18 months, we’ve spoken with hundreds of HubSpot customers about how they want to use AI at work. We heard two things.
First, it’s clear that growth-minded leaders recognize the value of AI, but current solutions fall short on delivering full value and growth potential for their businesses. Consumer AI tools are easy and fast, but disconnected from a business and its data. While enterprise tech solutions present the opposite challenge: they’re slow to implement and difficult to use. We took this as a call to action to develop a complete AI solution that delivers ease, speed, and unification.
Second, we heard that while AI needs to be easy to use, it also needs to be easy to trust. Businesses want to use AI to grow easier and faster while maintaining the trust they’ve built with their customers. That’s why our philosophy at HubSpot is that AI should be more than just a growth tool. It should empower your business while keeping security, privacy, and customer choice at the forefront.
How do you develop products for AI@Work?
Our AI ethics principles provide a framework to guide product development. This means we develop AI products that are easy, fast, and unified while also being compliant, trustworthy, and ethical. One recent example of this is the launch of our complete AI solution, Breeze. It’s embedded into each of our Hubs and across the platform to make it super easy for our customers to discover, adopt, and get consistent value from HubSpot’s AI features. From a trust perspective, Breeze provides safeguards for sensitive data, keeps humans in the loop, and our feature cards detail exactly how our AI interacts with customer data.
Additionally, we provide opt-out options, AI trust FAQs, and AI product-specific terms, and empower our internal teams to achieve responsible AI by design. Our ML engineers work closely with product design, AI Trust and Safety, Legal and Compliance teams, using standard AI libraries for the tech stack and adhering to privacy and security guidelines for AI product development.
What internal capabilities do people or companies need to harness the power of AI for their organization?
To harness the power of AI, establishing organizational governance is crucial for managing associated risks. In HubSpot’s AI adoption, we engage cross-functional teams to develop AI policies and guardrails that guide the responsible use and development of AI, conducting Security, Legal, and Compliance reviews to uphold strict data protection standards. When integrating ChatGPT, we developed an internal AI tool accessible for all employees for work purposes. The tool is built on OpenAI’s technology and harnessed with HubSpot’s privacy by design measures, including responsible use guidelines and feedback gathering channels. Additionally, we maintain internal guardrails for AI code generation, AI content creation, AI voice and image use, and data use in AI product development.
How has AI impacted the way you work with your team over the past year?
As the Legal and Compliance team, we serve as a strategic partner to the business, enabling the business to move fast on building AI while solving for privacy, security, and legal risks. We’ve been deeply engaged with AI initiatives, such as contracting AI vendors, disclosing legal terms for AI products, and developing an AI Ethics framework. We’ve tracked global regulatory changes, integrated governance into AI product development and internal use, and advised on the legal and ethical use of data in AI. We also collaborate cross-functionally to provide customers with transparency through AI feature cards, AI Trust FAQs, and fulfill AI opt-out requests.
What are some key takeaways for leaders using or planning to use AI@Work?
AI should be more than just a growth tool. It should empower your business while keeping security, privacy, and customer choice at the forefront. We’re thrilled to be part of the wider BSA community that’s advancing conversations around responsible AI. To learn more about HubSpot’s AI products and commitment to responsible AI use, please visit our Trust Center.
About the author:
Alyssa Harvey Dawson is Chief Legal Officer at HubSpot. She is responsible for overseeing the company’s legal and compliance teams. She has over 25 years of broad-based legal experience overseeing corporate governance, corporate, M&A and commercial transactions, enterprise risk management, compliance, data privacy, intellectual property, regulatory and governmental/public affairs.