AI technologies

mycode
Keeping it safe
AI technologies

Where we use AI technology we do it ethically and legally

This video has sub-titled versions in the following languages:
عربي, 中文, Nederlands, English, Français, Deutsch, Italiano, Español
Accessible via the 'cc' button in the video player menu once the video start

What it's all about

Technology development is fast moving and it is important that we remain up to date. A recent development is Artificial Intelligence (AI) and other information services that use AI technology.

The impact and implications of AI technology on Serco and the services we deliver are significant. Whilst AI technology can be used to automate many processes and tasks, we need to be aware of its potential impact in areas such as data privacy, and the ethical implications of using AI in public services.

Where we are considering or using AI technology we must ensure we do so ethically and responsibly, and ensure the data used does not breach any legal, regulatory, personal data privacy, client/supplier contractual requirements or the confidentially of Serco data.

What is AI?

AI covers a broad range of technologies, which can generally be grouped as follows:

Natural Language Processing (NLP): AI-powered tools that enable machines to understand and respond to human language. A large language model (LLM) such as GPT-3/4, is a type of machine learning model that can handle a wide range of natural language processing (NLP) use cases.
Generative AI: a category of AI algorithms that generate new outputs based on the data they have been trained on.
Computer Vision: AI-powered systems that can interpret and analyse visual data, such as images and videos.
Machine Learning: AI-powered algorithms that enable machines to learn and improve from data without being explicitly programmed.
Robotics: AI-powered machines that can perform tasks autonomously.
Predictive analytics: AI-powered tools that use data to predict future outcomes.

“Ethical and responsible use of AI and similar systems can support us in our commitment to impact a better future.”

What we all need to know and do

We only use data classified as Serco Public or for other data with written approval from the data owner (Divisional security for Serco Business classified data, or the data owner of client, customer or supplier data supported by a written assessment from Divisional Security Lead) when using AI or services that use AI.
We never submit data that contains Personal Identifiable Information (PII), such as names, addresses, CV’s, personnel records, health data, etc. or might breach third party Intellectual Property or put Serco Intellectual Property at risk.
We always use our Serco Email Address and set a complex password – See SMS Security & IT section for password guidance – when registering for a service but only after you have spoken with your Divisional IT team and had the service assessed for cyber and information security considerations.
We never use personal devices to access and use such AI technology services for Serco Business purposes.
We always check the output before disclosing it outside Serco, ensuring it is accurate, relevant and does not contain harmful, defamatory, biased, or copyrighted content and that it is reviewed for alignment with our standards and ethics before use.