Analysis
Tech In the City: The Cyber Security and TPRM Implications of Smart Cities
Cyber-attacks on organisations’ digital supply chains remain one of the greatest security threats facing organisations worldwide. Cybercriminals have become adept at identifying and exploiting weak nodes within supply chains to attack organisations upstream and downstream by exploiting their digital dependencies.
A 2025 report by Cybersecurity Ventures estimates that supply chain-related breaches have increased by 40% compared to 2023, costing companies billions in damages. According to the World Economic Forum’s latest Global Cybersecurity Outlook, due to the complexity of digital supply chains and the lack of visibility into supplier dependencies, supply chain vulnerabilities and risks are seen as the biggest barrier to greater cyber-resilience by 54% of large organisations.
The World Economic Forum expects the cyber-threat landscape in 2025 to be shaped by increasingly sophisticated attacks, with AI-powered cybercrime among the top concerns. While the widespread adoption of artificial intelligence (AI) across industries and applications brings significant efficiency, speed and performance benefits, it also offers the same powerful capabilities to cybercriminals looking to exploit supply chain vulnerabilities.
Tech-savvy hackers are having a field day deploying AI to amplify the sophistication, speed and accuracy of their attacks. AI can help attackers to very quickly identify weak points in supply chain networks and launch attacks at scale, with minimal effort.
Already in 2025 there have been high-profile supply chain breaches powered by AI tactics. SolarTrade, a logistics management platform used by more than 500 global retailers, was compromised when attackers used AI to inject malicious code into a routine software update. The breach gave hackers access to customer payment information, disrupting supply chain operations for months. In another attack, hackers used AI to target the firmware update system of Medtech, a major medical device manufacturer. Malware was inserted into life-saving devices, including pacemakers and insulin pumps, raising significant concerns about patient safety.
In the wrong hands, AI can be a devastating force, wreaking havoc across supply chains by exploiting vulnerabilities, evading cybersecurity defences and corrupting systems or exfiltrating data at speed and scale. Hackers are weaponising AI to attack organisations via their supply chains by various means.
Machine learning algorithms can be used to carry out rapid and accurate supply chain reconnaissance, quickly analysing huge datasets, such as vendor security policies or network traffic, to pinpoint vulnerable suppliers. These algorithms help attackers identify the weakest supply chain links, such as smaller suppliers with less robust security measures. AI enables all of this to be executed at a pace and scale not possible with conventional hacking techniques, like phishing or social engineering.
AI tools can be used to identify misconfigurations or outdated systems and exploit them during maintenance or software updates. AI-driven analytics can scan for these security lapses far more efficiently than manual reconnaissance. AI can also be used to manipulate application programming interfaces (APIs) and other external interfaces to attack all connected systems.
That means any business using mainstream AI solutions or integrating AI into their tech stacks via APIs is vulnerable to exploitation by hackers. Cybercriminals use AI to target the open-source datasets used by AI software, and corrupt them with misinformation or malicious code. Once that compromised code enters the AI model training process, it creates a ripple effect that corrupts the entire AI software ecosystem.
Attackers are using AI to develop corrupting software updates or evasive malware that can bypass traditional detection methods. AI-generated malware is self-evolving, which means it can dynamically learn from its surroundings and make autonomous decisions about its strategy, based on the data it gathers. Unlike conventional malware, which follows static, pre-defined rules, AI-powered malware can intelligently infiltrate systems and sustain attacks while evading detection.
Once inside an organisation’s network, AI can explore laterally across systems to detect valuable data or further vulnerabilities. It can then extract data without detection, by encrypting the data to avoid triggering security alarms. Finally, AI can even erase the evidence of an attack, making it much harder for security teams to identify the breach and trace its origin.
Malicious software can quickly spread throughout supply chains. Once a single vendor is compromised, AI-powered malware can move freely into connected systems and cascade throughout supply networks, impacting multiple downstream organisations.
Digitalisation has transformed the world of supply chains. The proliferation of digital devices and cloud-computing systems has enabled organisations to connect easily with suppliers, partners or clients anywhere in the world. Outsourcing has become simpler and more cost effective than ever, with many organisations now routinely outsourcing everything from payroll and pension services to payment platforms and customer support functions – creating a complex web of supply chain relationships.
In digital supply chains, organisations are linked to many direct suppliers as well as many hundreds of indirect connections such as subcontractors or suppliers’ suppliers and so forth. With such a vast matrix of connected suppliers, vendors and service providers, a breach anywhere in the network could quickly impact organisations throughout the whole ecosystem. The range of internet-connected systems and devices businesses rely on every day now provides a huge – and ever-growing – attack surface for hackers to exploit.
Many suppliers and vendors, as well as their clients, have now integrated AI into their systems and processes to deliver efficiencies and improve performance. This includes third-party software providers who are using AI to enhance their own services. This introduces potential new risks into supply chains. Without the necessary expertise or risk-management systems in place, it’s difficult for organisations to evaluate the extent of AI-related security risks in their supply networks.
As traditional cybersecurity defences struggle to keep pace with AI-powered threats, AI itself can play a critical role in helping organisations fight back. It can be used in a variety of ways to help companies detect and prevent fast-evolving AI-driven attacks.
AI can be used to constantly monitor networks for threats and anomalies in real time. Traditionally, malware defences have been reactive, analysing systems for known threats and then blocking or removing them. AI enables organisations to take a more proactive approach, continuously tracking network activity, learning and adapting to emerging threats.
AI systems can automatically identify malicious activity within networks or systems, as well as identifying anomalies that could signify an attack. By constantly analysing huge volumes of data, AI can identify patterns and any deviations that may be a cause for concern. Machine learning algorithms can also be used to detect zero-day vulnerabilities in systems and software, using predictive analytics to anticipate these vulnerabilities before they can be exploited by hackers.
As organisations across multiple sectors recognise the risks associated with supply chains, improving supplier risk management has become a top priority. AI tools are now being used to conduct continuous risk assessments of supplier security postures and performance, as well as monitoring market conditions and other external factors that could influence supply chain risks. AI can help with checking, or validating, supplier compliance with risk-management frameworks and to ensure continued compliance over time.
Suppliers themselves can use AI tools to simplify and enhance the completion of security questionnaires, using automation to extract relevant information and proof from key policies and security certificates, and ensuring the latest and most comprehensive data is used to answer all questions.
If a cyber-attack, data breach or other security incident is detected, AI can support the rapid containment, response and recovery – to minimise financial and reputational impacts. Organisations should have incident-response plans in place, setting out the system containment, customer notification and other processes to be enacted in the event of a breach. AI can accelerate the response to an incident by automatically triggering these pre-planned workflows and actions. By doing so, AI-driven response systems can help to minimise the damage, isolate affected systems and prevent malware from spreading further.
Many organisations have already implemented zero-trust security architecture as a way to combat growing cybersecurity threats. Zero-trust security solutions consider all users and devices to be potential security threats, and use continuous verification and strict access controls to prevent unauthorised entry.
AI tools can now be used to strengthen zero-trust security and make it even more resilient, by automating, accelerating and improving the accuracy of verification processes. Zero-trust frameworks can also be strengthened by real-time AI threat identification, helping to detect AI-driven attacks that seek to thwart zero-trust security measures.
The nature of AI, with its ability to continuously learn and adapt, means that the threats it poses to supply chains will continuously evolve. At the same time, those same capabilities will enable enhanced supply chain defence and risk-management processes to evolve in parallel.
Advances in computing are likely to strengthen the powers of AI and exacerbate the risks. The arrival of quantum computing, with its exponentially greater problem-solving capabilities, could transform computational power in ways that are not yet fully understood. The integration of machine learning with quantum computing could pose unfathomable risks to supply chain security – as well as potentially generating new tools to combat those risks.
Supply chain risk management practices in future will also be shaped by new digital resilience and cybersecurity regulations, such as the Digital Operational Resilience Act (DORA) and Network and Information Systems Directive 2 (NIS2). These regulations require organisations to monitor and manage suppliers and vendors across complex digital supply chains – a process that could be significantly enhanced by the application of AI.
The increasing sophistication and ubiquity of AI-powered supply chain attacks mean that organisations need to step up security and risk management across their supply chains. Harnessing AI can supercharge those efforts, from threat detection and response to zero-trust security enhancement and automated risk assessment. Organisations need to weigh up the costs and benefits of AI implementation, ensuring solutions are proportionate to the risks and applied rigorously across ever-growing supply networks.
Sign up to our monthly newsletter to receive exclusive research and analyses by our experts, the latest case studies from our clients as well as guides, explainers and more to turn your supply chain risk management programme into a resounding success story.
