The News: Twilio debuted its CustomerAI Privacy Ladder and AI Nutrition Facts labels at its SIGNAL 2023 user conference, each of which are designed to provide more transparency and accountability for customers using the company’s AI technology. The CustomerAI Privacy Ladder is designed to classify AI models based on three levels of data privacy and complexity, while the AI Nutrition Facts labels are designed to provide transparency around the type of model used, what data is used in the model, how the data is used, and compliance information, in a concise, easy-to-digest format.
Analyst Take: Twilio made a host of interesting product announcements based around the introduction of CustomerAI, a set of capabilities woven throughout all its products that couple the power of large language models (LLMs) with real-time customer data flowing through Twilio’s Customer Engagement Platform. Along with the product announcements, Twilio focused on its commitment to using generative and predictive AI by providing details on its CustomerAI Privacy Ladder, and rolling out its AI Nutrition Facts Labels, which provide transparency around the AI models and the data that power them.
Customer AI Privacy Ladder Offers Insight Into Data Use and Privacy Levels
The CustomerAI Privacy Ladder is designed to graphically illustrate how Twilio classifies the AI models it uses, based on the type of data being used, the intended user(s), and whether personally identifiable information (PII) is included or removed from the model.
Level 1 models include company-specific data, without PII, and are only geared for that company’s specific use. Because there is no PII incorporated, and the model is only deployed for a single company’s use, the data privacy risk is relatively low. Level 2 models incorporate PII, which raises both the risk level, but should also improve personalization, contextual features, and accuracy. Because PII is incorporated, Level 2 models are only deployed within a single company.
Level 3 models incorporate company data, without PII, and are designed to be used across multiple customers. Level 3 models that incorporate customer data can be more accurate, as they are trained on a wider range of data, but due to privacy concerns, PII will always be redacted.
This transparency will help organizations decide which models are most appropriate, based upon the level of functionality and risk tolerance they are willing to accept. While the CustomerAI Privacy Ladder is a welcome tool for organizations seeking more transparency, Twilio executives mentioned that it is likely a work in progress and may be further refined and expanded upon over time.
AI Nutrition Facts Provide Transparency and Insights into AI Model Design
Twilio also announced its AI Nutrition Facts Labels, a consistent and transparent tool to help customers understand exactly “what’s in the box” when assessing Twilio’s CustomerAI products and features across the platform.
The AI Nutrition Facts Labels provide key information about an AI model’s privacy level and design elements in a format the mimics food nutrition labels, providing critical details including the CustomerAI Privacy Ladder Level, the model type, which base models and owners are involved, and a detailed breakdown for how data is used by each model, and any relevant links to supporting, public-facing resources.
In an environment where most vendors are talking about model and data transparency, it is refreshing to see Twilio take the lead by providing detailed information about the models they use in a relatively simple-to-understand format. The “Trust Ingredients” provide a fairly granular level of detail while not being overwhelming to non-data scientists, and the company has also launched nutrition-facts.ai, a new AI Nutrition Facts Label generator that is powered by Twilio, allowing its customers and partners to easily and thoughtfully create their own AI Nutrition Facts Labels.
Transparency and Accountability Will Help Twilio Walk the Walk With Responsible AI
In my conversations with customers, many are excited about the prospect of deploying generative AI tools and functionality to improve efficiency and unlock new ways of interacting with data. But data privacy, data security, and the fear of having private data unwittingly being fed back into a public model is still very much a key concern.
Twilio’s two-pronged approach of delineating its AI offerings into Levels, and then providing a transparent accounting of the type of model used via its AI Nutrition Fact Labels, is a good step in the right direction. It provides a framework or starting point for discussions between Twilio and its customers about how AI is being used within the company’s platform, and illustrates that Twilio is not simply issuing blanket statements about responsible AI with little detail.
It will be interesting to see whether other platform vendors follow Twilio’s example, which likely would help to make customers, partners, and regulators more comfortable as additional AI use cases are deployed.
Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.
Other insights from The Futurum Group:
Solid Q2 2023 Results for Twilio With 10% Total Revenue Growth YoY
Twilio and AWS Expand Partnership, Leveraging Predictive AI
Improving Contact Center Experiences via NLP, NLU, and Analytics-Focused AI
Author Information
Keith Kirkpatrick is Research Director, Enterprise Software & Digital Workflows for The Futurum Group. Keith has over 25 years of experience in research, marketing, and consulting-based fields.
He has authored in-depth reports and market forecast studies covering artificial intelligence, biometrics, data analytics, robotics, high performance computing, and quantum computing, with a specific focus on the use of these technologies within large enterprise organizations and SMBs. He has also established strong working relationships with the international technology vendor community and is a frequent speaker at industry conferences and events.
In his career as a financial and technology journalist he has written for national and trade publications, including BusinessWeek, CNBC.com, Investment Dealers’ Digest, The Red Herring, The Communications of the ACM, and Mobile Computing & Communications, among others.
He is a member of the Association of Independent Information Professionals (AIIP).
Keith holds dual Bachelor of Arts degrees in Magazine Journalism and Sociology from Syracuse University.
