Overview
Heading is built on artificial intelligence. Our platform tracks the answers given by AI assistants and AI search engines, analyses those answers, and uses AI to help you understand what those answers mean for your brand and what to do next.
This AI Policy explains how we use AI, what kinds of third-party providers are involved, what happens to your data, and the limits of what AI can tell you. It should be read together with our Terms of Service and Privacy Policy. Capitalised terms have the meaning given in our Terms of Service.
This policy is published by Heading AI Limited, a company registered in England and Wales with company number 17241813.
How we use AI
Querying AI platforms. The core of the Services is running the prompts in your portfolio against Third-Party Platforms, such as leading AI assistants and AI search engines, and recording the answers, the brands mentioned and the sources cited. The platforms we track will expand as the product develops. We collect these answers on your behalf, either directly or through specialist third-party data providers.
Analysing answers. We use AI models to read the answers we collect and identify which brands appear, whether a source is cited or merely mentioned, what sentiment and qualities are associated with a brand, and how answers change over time. This analysis produces the visibility, citation, competitor and brand perception data you see in the platform.
Prompt research. We use AI, together with search demand data, to suggest the questions buyers are likely to ask in your category and to help you build a prompt portfolio worth tracking.
Crawler and traffic analysis. We use rules and models to classify crawler requests in your server logs by operator and purpose (retrieval, search or training) and to attribute website sessions to AI platforms in your analytics data.
Heading Agent. The Heading Agent is a conversational assistant that can investigate your data, explain a change, compare competitors and draft tasks and recommendations for your team to review. It is powered by third-party large language models and has access to the data in your workspace.
Reporting. We use AI to draft summaries, findings and next steps for stakeholder reports, which you can edit before sharing.
Which AI providers we use
Separately from the Third-Party Platforms we track, we use commercial AI models and APIs from third-party providers to power our own analysis and agent features. We access these models directly or through an AI gateway service that routes requests to the relevant model provider. The AI providers and gateway services we currently use are listed in our sub-processor list at docs.useheading.com/legal/sub-processors. We may add, remove or change providers and models as the market develops, and we will keep our sub-processor list up to date.
We choose providers whose terms allow business use of their APIs without training on customer data, and we use them (whether directly or through an AI gateway) on that basis.
What happens to your data
Your data is used for your account. Your prompts, brand data, connected data and agent conversations are used to provide, secure, support and improve the Services. We do not share them with other customers, and we only use them for the benefit of anyone else in aggregated and anonymised form, as described in our Terms of Service.
We do not train AI models on your data. We do not use your Customer Data, agent conversations or Outputs to train or fine-tune general-purpose AI models, including large language models.
Our providers do not train on your data. The terms on which we use our AI providers do not permit them to use the content we send them to train their models. Where a provider offers zero data retention or short retention periods for API requests, we use those settings where available.
We send only what is needed. When we use an AI model to analyse an answer or respond to an agent question, we send the relevant content and context for that task. We do not send your account credentials, payment details or unrelated workspace data.
Prompts sent to AI platforms are about brands, not people. The prompts we run against Third-Party Platforms describe products, categories, markets and brands. They do not contain your account details, and you should not include personal data about individuals in your prompt portfolio.
Answers may mention people. AI answers and the web pages they cite may incidentally refer to named individuals, such as founders, authors or reviewers. We store that content as evidence for your review and do not use it to build profiles of individuals.
Data is stored securely. Data is encrypted in transit and at rest, and access is limited to the people and systems that need it. Our Privacy Policy and Data Processing Terms describe our security measures in more detail.
Accuracy and limitations
AI answers are not stable. The same prompt can return different answers on different days, in different locations, on different models, or for users with different histories. What we record is a snapshot of what a platform returned to us at a particular time, and it may not match what an individual user would see.
Visibility data is an observation, not a guarantee. A brand appearing in an answer does not mean a buyer saw it. A crawler requesting a page does not mean the page was cited. A successful crawl does not mean a sale. We design the platform to keep these things separate so that you can inspect the evidence behind a number.
AI analysis can be wrong. Models can misidentify a brand, misjudge sentiment, mis-attribute a citation or produce a summary that is plausible but incorrect. We test and monitor the quality of our analysis and correct errors where we find them, but we cannot guarantee that every classification is accurate.
The Heading Agent can make mistakes. Its explanations and recommendations are generated from the data in your workspace and from general models. They are a starting point for investigation and decision-making, not a substitute for your own judgement or for professional advice. Please review what the agent tells you, and any tasks or content it drafts, before acting on it.
We cannot guarantee that AI will recommend your brand. AI platforms decide what to include in their answers, and their behaviour changes. Heading helps you understand where you stand and what to try. It does not control what any AI platform says.
Human oversight and transparency
AI is used throughout the platform, and much of what you see, including analysis, agent responses and drafted summaries, is generated or assisted by AI. We do not label every AI-assisted output individually, but this policy explains where and how AI is used so that you know what you are looking at.
Everything the platform drafts for you, including tasks, findings and report text, can be reviewed and edited by you before it is shared with anyone.
The original evidence behind a finding, including the answer text and the sources it cited, is kept in the platform so that you can check it.
Our team reviews the performance of our AI features regularly and investigates reports of incorrect or unexpected output. If you spot an error, please tell us at support@useheading.com.
Responsible use
We are committed to using AI responsibly. In particular, we:
are transparent about where and how AI is used in our product;
keep your data private and secure, and do not use it to train general-purpose AI models;
design our features to help you inspect evidence rather than rely on a single score;
do not use AI to profile or track individuals, and do not permit our customers to do so through the Services;
do not use the Services to generate spam, deceptive content or content designed to manipulate AI platforms or their users;
respect the terms of the AI platforms we work with and adjust our methods when those terms change; and
keep this policy under review as the technology and the law develop, including any requirements under the EU AI Act and UK guidance that apply to our Services.
Your responsibilities
When using AI features in the Services, you agree to:
review Outputs before relying on them, and take responsibility for decisions you make based on them;
avoid entering personal data about individuals, confidential third-party information or special category data into prompts, documents or agent conversations unless you have the right to do so;
comply with any rules in your jurisdiction about disclosing or labelling AI-generated content that you publish; and
use the AI features in line with our Terms of Service and acceptable use rules.
Changes to this policy
We will update this policy as our use of AI develops. We will post changes on our website and update the date at the top. Where a change materially affects how we use your data, we will let you know through the Services or by email.
Contact
If you have questions about how Heading uses AI, please contact us at support@useheading.com, or write to Heading AI Limited, 124-128 City Road, London, England, EC1V 2NX.