How Better Anonymisation Unlocks UK National Security Innovation
Dr Hayley Watson
Director, Innovation & Research, Trilateral Research
Dr Joshua Hughes
Cluster Lead - Law Enforcement and Community Safeguarding, Trilateral Research
As the UK prepares to launch its National Security Strategy in June 2025, there's an urgent focus on innovation and growth. Such ambitions require a significant step-change in securely and efficiently processing sensitive data compared to existing slow manual methods, which is vital for operations including investigations and intelligence among others.
Why Data Anonymisation Matters for National Security
Data anonymisation is the process of removing or altering direct and indirect personal identifiers to prevent the identification of individuals. It plays a pivotal role in balancing the utilisation of data for national security purposes with the protection of individual privacy rights. Effective anonymisation enables the sharing and analysis of data across agencies and with trusted partners without compromising personal privacy. This is particularly important in national security contexts, where access to comprehensive datasets can enhance threat detection, intelligence capabilities, and criminal justice capabilities.
Current Issues with Anonymisation:
UK public-sector anonymisation methods currently face several issues:
Slow Manual Processes: Some UK public-sector agencies still rely heavily on manual redaction techniques. These tedious processes involving manually obscuring sensitive information from documents. These outdated methods can delay operations by days or even weeks, slowing down vital decisions in situations like law enforcement investigations or child protection cases.
Risk of Human Error: Manual anonymisation carries inherent risks of human error. A single oversight can result in a serious data breach, exposing sensitive information and compromising operational security, public trust, and organisational credibility. An analysis of ICO data found 39% of all data breaches were due to human error.
GDPR and Legal Compliance Risks: Organisations handling personal data face rigorous GDPR requirements. Manual or inadequate anonymisation can leave organisations exposed to significant compliance risks and potential fines from regulatory bodies like the Information Commissioner’s Office (ICO).
Inefficient Collaboration: Lack of robust anonymisation tools hampers collaboration between UK public-sector bodies, such as law enforcement, NGOs, and healthcare agencies. Inefficient data sharing significantly impacts joint operations, such as coordinated responses to national security threats or cross-agency safeguarding efforts.
Challenges Specific to the UK Context: In particular, child safeguarding and police
collaboration with NGOs highlight the urgent need for improved data-sharing practices. UK safeguarding reviews have repeatedly emphasised that slow data exchange and ineffective anonymisation processes directly affect frontline service delivery and the protection of vulnerable groups.
The Alternative: Technology-Driven Solutions
Given these significant challenges, traditional manual methods are clearly insufficient. To overcome these barriers, the national security sector urgently requires innovative, technology-driven solutions that guarantee rapid, reliable, and secure anonymisation.
Trilateral Research’s MASC is specifically designed to meet this need. MASC combines cutting-edge technology with human oversight, resolving today’s anonymisation challenges and paving the way for secure, data-driven innovation.
Introducing MASC: The Future of Secure Data Sharing & Innovation
MASC leverages state-of-the-art Natural Language Processing (NLP) to effectively detect and anonymise both direct (like names) and indirect personal identifiers (like locations or potentially unique events) in unstructured text. Effective anonymisation aims to achieve a level of data protection that meets or exceeds the legal threshold in data protection legislation. Key features include:
Human-in-the-loop Interface: Users retain control, reviewing and refining anonymisation decisions, ensuring accuracy and explainability and balancing privacy and utility by preserving context.
Accessibility and Inclusivity: Built to WCAG 2.2 standards, MASC ensures usability across diverse operational contexts.
Bias and Audit Checks: Embedded bias-detection tools align MASC with the MOD’s dependable AI standards, helping maintain fairness and transparency.
In recent trials, MASC identified personal data with 94–99% accuracy, nearly double the accuracy of leading commercial NLP services. Additionally, MASC processed 2,000 case files in just 13 minutes, a task that previously required analysts six days, representing significant operational efficiency gains.
Our expert researchers have reviewed relevant legal, technical, and behavioural aspects to develop a framework for understanding the quality and resilience of anonymisation techniques to determine whether outputs of MASC can be seen as reliably anonymous under data protection law. Our evaluations confirm MASC achieves anonymisation above the (UK) GDPR-defined threshold in up to 95% of test documents comprising sensitive survivor testimonies.
MASC supports various aspects of UK national security, such as:
Enabling secure disclosure for prosecution
Secure intelligence sharing
Management of data in high-stakes investigations
AI Innovation – supporting model training and responsible innovation
As the National Security Strategy launch approaches, adopting robust, technology-driven anonymisation solutions like MASC is essential for the UK's future security, growth, and innovation.
To explore partnership opportunities and trial MASC, contact our Research Services team at ResearchServices@trilateralresearch.com.
