In security applications, how can we rely on AI to show us all the possible threats (not seeing what we are not shown)?

Background

Increasing our global impact recognises the UK’s opportunities in the world and the importance of DfT contributing effectively toward the government’s vision of a global Britain. Operating and engaging globally will play a vital role in assisting the department in maintaining and extending its influence with international partners, including ensuring high standards of transport safety, security, and environmental protection and boosting UK trade, exports, and inward investment. UK science and innovation has the potential to develop transport-related exports and inward investment, thereby helping grow and level up the economy. Our R&D around climate change mitigation and adaptation can help influence the international community to take progressively stronger action.

Next steps

Get in touch with bridgetoresearch@dft.gov.uk

Source

This question was published as part of the set of ARIs in this document:

DFT-Areas of research interest 2023 GOV UK

Related UKRI funded projects

  • Cyber Graph-to-Text: AI automation for Threat Intelligence, made accessible to all

    The intelligence cycle is the process of collecting, analysing, and disseminating intelligence to help decision makers understand their environment and make appropriate decisions. However, the human mind is not capable o...

    Funded by: Innovate UK

    Lead research organisation: ELEMENDAR LTD

    Why might this be relevant?

    The project focuses on using AI to automatically identify, quantify, and predict cyber-threats, addressing the question of relying on AI to show all possible threats.

  • ATM: Automated Threat Modelling for Enterprise AI-enabled Assets

    In today's AI era, most companies use AI assets incorporating machine learning and deep learning models. In this context, AI assists enterprises in their decision-making process. The estimated cost of building and implem...

    Funded by: ISCF

    Lead research organisation: UNIVERSITY OF SHEFFIELD

    Why might this be relevant?

    The project proposes an Automated Threat Modelling system to detect all threats related to AI-enabled assets, addressing the question directly.

  • AISecure

    The AI market size is expected to be USD 190.61 billion by 2025, with a CAGR of 36.62% between 2018 and 2025. Furthermore, AI systems will be used in a wide range of applications, from cybersecurity threat prevention to ...

    Funded by: Innovate UK

    Lead research organisation: ROYAL HOLLOWAY UNIV OF LONDON

    Why might this be relevant?

    The project addresses the question fully and the authors have the necessary expertise in AI systems and cybersecurity.

  • Academic Centre of Excellence in Cyber Security Research - Cardiff University

    Cardiff University is a leading UK academic institution for cyber security analytics and artificial intelligence - the interpretation and effective communication of applied data science and AI methods through interdiscip...

    Funded by: EPSRC

    Lead research organisation: CARDIFF UNIVERSITY

    Why might this be relevant?

    The project partially addresses the question by focusing on cyber security analytics and artificial intelligence, but does not specifically mention reliance on AI to show all possible threats.

  • Vouchsec - Conversational XDR Platform for the AI Era

    Vouchsec offers a solution that shortens cyber threat processing time tenfold and detects emerging AI-powered attacks. The main target market are security operation centres (SOCs), that can be either in-house company div...

    Funded by: ISCF

    Lead research organisation: UNIVERSITY OF OXFORD

  • Machine Learning, Robust Optimisation, and Verification: Creating Synergistic Capabilities in Cybersecurity Research

    The need for better support to deal with the threats of cybersecurity is undisputed. Organisations are faced with an ever growing number of malware and integrated malware attack tools, attempted attacks on infrastructure...

    Funded by: EPSRC

    Lead research organisation: Imperial College London

    Why might this be relevant?

    The project partially addresses the question by exploring new modelling and analysis capabilities for cybersecurity, but does not specifically mention reliance on AI to show all possible threats.

  • Bayesian Analysis of Competing Cyber Hypotheses

    Cyber security is recognised as important at the highest levels of international government. President Obama has said that "the Cyber threat is one of the most serious economic and national security challenges [the ...

    Funded by: EPSRC

    Lead research organisation: University of Liverpool

    Why might this be relevant?

    The project focuses on using AI to analyze cyber threats and proposes a technique to improve decision-making in cyber security operations.

  • UKRI Trustworthy Autonomous Systems Node in Security

    Autonomous Systems (AS) are cyberphysical complex systems that combine artificial intelligence with multi-layer operations. Security for dynamic and networked ASs has to develop new methods to address an uncertain and sh...

    Funded by: SPF

    Lead research organisation: Lancaster University

  • ATM: Automated Threat Modelling for Enterprise AI-enabled Assets

    In today's AI and Digital era, most companies work with, deploy, or are using AI assets incorporating machine learning/deep learning models. In this context, AI assists enterprises in their decision-making process. Howev...

    Funded by: ISCF

    Lead research organisation: UNIVERSITY OF SHEFFIELD

  • Speak-to-Secure (S2S): A Natural Language Human-Computer Interaction To Secure Cyber Physical Systems

    Securing cyber physical systems requires multi-domain expertise. With the current cyber security workforce gap, it becomes harder to find this multi-domain skillset. A key entry barrier is learning a large number of tool...

    Funded by: ISCF

    Lead research organisation: DE MONTFORT UNIVERSITY

    Why might this be relevant?

    The project addresses securing cyber physical systems using AI-powered technology for natural language interaction, which partially relates to the question.

Similar ARIs from other organisations

Cyber Security and Digital Identity (CSDI) - Cyber security of technologies and services Does deploying AI models in an existing process or system increase the attack surface of the host system? If so, how can this be mitigated? Who would be the owners for securing against those vulnerabilities and, where this is unclear, should organisations be sharing knowledge and security ownership, across the technology stack and system of deployment? Analyse and examine the full extent and range of cyber security risks to the UK economy in software development, distribution and supply chains. Identify what risks to the UK’s cyber resilience are introduced through software supply chains by software vendors and users? Which are the highest profile risks? We are interested in B2B/enterprise software rather than consumer software but including: operational technology, IT and cloud and on-premise. Department for Science, Innovation & Technology, 2024
See details
Measuring and responding to security threats faced by the UK How do we make risk-reduction real at the local level where it is not realistic to extrapolate from large-scale datasets and statistics? Cabinet Office, 2019
See details