Introduction
In today's ever-evolving world, artificial intelligence (AI) is quickly becoming an integral part of our daily lives. From personalised music and movie recommendations on our favourite streaming services, to seamless customer service interactions through AI-powered chatbots, AI is revolutionising industries across the board. At SSEN Transmission, we recognise the potential of AI to enhance our operations and improve customer experiences.

We understand the importance of utilising cutting-edge technologies to not only keep our costs low for GB energy consumers but also to drive positive societal impacts. That's why we're excited to embark on our journey of integrating AI into our business, particularly in the analysis of stakeholder and customer feedback.

Through transparent and responsible AI implementation, we aim to deliver quicker analysis of feedback, allowing us to take action promptly. Whilst we’re leveraging cutting-edge AI technology to analyse stakeholder and customer feedback, we’re proceeding cautiously given the newness of this technology, ensuring a gradual integration that complements and enhances the efforts of our dedicated teams. This approach allows us to combine the strengths of AI-driven insights with the invaluable expertise and human intelligence of our people. We’d like to reassure stakeholders and customers that all data processed by our new AI tool will be kept within SSEN Transmission.

We have provided useful information below which explains how we’re currently utilising AI in SSEN Transmission (please click the links below):


Data Collection:
We collect feedback on our projects, proposals, engagement, information provided, and service from stakeholders. We collect feedback from you in different ways: the webform, emails, letters and in person. We would ask that you ONLY provide detail about your opinion of the particular project that you are commenting on and avoid introducing anything that could identify you (personal data), where possible (unless it relates to your particular concern or wish us to contact you). For example, if you give us your contact information such as names, addresses, and contact details, we will process these. Any health or other personal data provided voluntarily by you will also be collected, so we can take note of these points and take appropriate action that may be required. If you do not wish this information to be processed, please avoid introducing any of your personal data in your response.

How does the AI analyse the feedback?
The AI has been developed to recognise key words and categories (these controls and instructions have been developed by humans) and to recognise tones and themes. The AI Tool will perform the analysis and provide a summary of the key things raised from the feedback. This summary will then be reviewed by humans to understand what the overall response to the feedback is and to recognise trends in the comments.

How are we ensuring that the AI analysis is correct, free from bias or discrimination?
Ensuring the accuracy and fairness of AI analysis is paramount at SSEN Transmission. We employ a rigorous process to mitigate bias and discrimination in our AI-driven feedback analysis.

Firstly, we conducted cross-check activities where anonymous feedback was reviewed manually alongside AI analysis of the same data. By comparing results, we are able to identify discrepancies and fine-tune our AI to improve accuracy. This is an activity we will undertake on an ongoing basis to ensure our AI Tool remains fit for purpose.

Secondly, we are proud to utilise the state-of-the-art AI engine, GPT 4.0, renowned for its advanced language model capabilities. We upgraded from GPT 3.5 to 4.0 which has provided us with improved accuracy and understanding of the feedback provided.

Since GPT 4.0 is developed by a third-party organisation, they are responsible for training the AI and therefore it's not influenced by SSEN Transmission’s internal biases or agendas. Categories are provided to our AI tool to help it understand the environment it is operating in, for example, who we are as an organisation, that our stakeholders include local communities impacted by our infrastructure projects across Scotland, what our proposed projects are, and environmental aspects including visual amenity, wildlife, and many others.

We understand concerns about spelling mistakes potentially impacting the performance of AI tools. We would like to assure stakeholders that our AI tool is designed to handle variations in spelling and still provide accurate analysis.

GPT 4.0, the AI model powering our tool, has been trained on a vast amount of text data by OpenAI, including variations in spelling, syntax, and grammar. This enables it to understand and interpret text input with a high degree of accuracy, even in the presence of spelling mistakes. GPT 4.0 excels at understanding the context of a given text, which helps it make sense of misspelled words within the broader context of the sentence or paragraph. By considering surrounding words and phrases, the AI tool can understand the meaning, even if there are spelling errors present.

Despite advancements in AI technology, we recognise the importance of human review. That's why, we always conduct manual checks to validate the accuracy of our AI analysis. This approach ensures that our AI tools are not only efficient but also trustworthy, instilling confidence in the insights they provide.

Is the AI analysis reviewed by humans?
Yes, we will be testing a percentage of the AI analysis by having a human review the analysis against the original piece of feedback to check for accuracy.

Is my response or personal data being used to train the AI tool?
No, we used a Microsoft product to develop this AI and this does not use the data to train itself. This tool is a private tenant AI (meaning it stays within our IT systems and information is not shared outside of SSEN Transmission). We'd like to reassure our stakeholders that any data that may be processed at a future date by our new Al tool, will not be introduced into any public AI (such as Chat GPT), or used to train any AI model.

What governance and assurance was followed when developing the AI tool?
To ensure our AI tool was developed ethically and is producing accurate analysis, our project team has collaborated with our Data Assurance and Legal experts to adopt best practice and follow the Information Commissioner’s Office (ICO) guidance on implementing AI. We would like to reassure stakeholders that we have conducted validation testing before implementing our AI tool to assist us in analysing stakeholder feedback. Our project team responsible for our AI tool continues to work collaboratively with our Assurance and Legal teams on an ongoing basis to ensure the appropriate governance is in place. 

Any changes to ICO guidance or GDPR requirements will be reviewed by our assurance and legal teams, who will then feed these updates into our AI project team. The project team will incorporate them into the AI system to ensure ongoing compliance with the latest regulations, maintaining robust data privacy and ethical standards.

Purpose and Legal Basis
The purpose of collecting this data is to understand stakeholder concerns regarding our impactful infrastructure projects. This understanding aids us in mitigating these concerns to support successful planning consents and positive engagement with impacted communities. The legal basis for this data collection is to fulfil our legitimate interests in improving our services and engaging effectively with stakeholders. Regarding our legal basis for processing any special category data such as details about your health, we rely on substantial public interests, protecting children and adults at risk and in defence of or establishment of legal claims.

Who do we share with?
Personal data contained in your feedback processed by our AI tool will remain within SSEN Transmission’s IT estate (unless we are required to share for legal or regulatory reasons). Key themes or trend analysis results determined by our AI tool may be shared with statutory consultees, other SSE companies, or regulators (however you will not be identifiable from these results).

Will the AI be making Automated Decisions or Profiling?
While the intention of the AI is to decide or match categories or themes to feedback to provide an overall snapshot, which will later support our decisions on the overall feedback, we will not make automated decisions about customers or stakeholders using our AI tool. Its purpose is to expedite the categorisation of feedback received, so we can deliver quicker analysis and allowing us to take action promptly.

How long will you keep my personal data?
We will only retain your personal data and feedback for as long as reasonably required. The feedback collected will be retained for the duration of the development of an infrastructure project plus an additional two years (unless we are required to increase the retention period for legal and regulatory reasons). This ensures that we maintain visibility of concerns raised throughout the lifetime of a project and can continue to address enquiries for a reasonable period post-energisation. For example, if you provide feedback on a proposed project due to be delivered by 2031, your data would be retained in our AI tool until 2033 (2031 + 2 years).

For more information on how we handle and process your data, please visit our Privacy Notice.

Can I object to processing?
You have the right to object to certain types of processing in certain circumstances. In particular, the right to object to the processing of your personal data based on our legitimate interests or on public interest grounds. If you object to your data being processed by AI, please contact our Data Protection team at: TransmissionDataProtection@sse.com

For more information on your rights or if you would like to exercise any of your rights outlined within our Privacy Notice, you are welcome to get in touch with our Data Protection team. 

This page was last updated on 30 September 2024.