Hi SupTech Community,
Welcome to the 49th issue of the Cambridge SupTech Lab bi-weekly LinkedIn newsletter, a source for updates on recent innovations, breakthroughs, opportunities, upcoming events in the suptech field, and news from the Cambridge SupTech Lab team.
If you would like to flag any items for inclusion in the next newsletter, please email us at cambridgesuptechlab@jbs.cam.ac.uk. Occasionally, we email our global community of supervisors, data scientists, vendors, and suptech experts to share event invites, news, or new courses—subscribe here.
This edition includes news from the Bank for International Settlements, Bank of England, Bangladesh Securities and Exchange Commission, European Banking Authority, Financial Conduct Authority, Financial Stability Board, Hong Kong Monetary Authority, Hungarian National Bank, International Monetary Fund, Monetary Authority of Singapore, Reserve Bank of India, Swiss Financial Market Supervisory Authority, and others.
Suptech Innovations
Hungary’s Central Bank (MNB) launches an artificial intelligence (AI)-powered fraud detection system for bank transfers. The MNB has initiated the first phase of its Central Fraud Detection System (CFDS), an AI-driven platform designed to swiftly identify fraudulent bank transfers. This cutting-edge system, part of the MNB’s broader “five strikes” anti-cyber fraud initiative, requires domestic banks to submit instant transfer data from July 2025, enabling enhanced risk assessment and fraud prevention. The MNB is also introducing stricter regulations to hold banks accountable for fraud losses and launching a public campaign to raise awareness of financial scams. Read more here.
Central Banking & Technology
Bank Negara Malaysia (BNM) and Payments Network Malaysia (PayNet) set to roll out an AI fraud detection system in 2026. In response to the sharp rise in online scams, the BNM and PayNet are developing an AI-powered system to monitor and flag suspicious bank transfers in real time. Using machine learning, large language models, and predictive analytics, it will alert users to potential fraud before transactions are completed. The initiative forms part of BNM’s wider digitalisation push, which includes the Digital Asset Innovation Hub unveiled in June 2025 and the National Fraud Portal launched in 2024. PayNet is also building “FinancialGPT,” a federated AI model trained on local financial data to enhance fraud detection while maintaining data privacy. Read more here.
Bank of England (BoE) outlines approach to responsible innovation in AI, distributed ledger technology (DLT), and quantum computing. The BoE has published its approach to fostering responsible innovation in AI, DLT, and quantum computing, technologies it sees as transformative for the UK economy. The Bank aims to create a robust environment that supports safe adoption of these innovations while safeguarding financial stability and promoting productivity. Emphasising collaboration with industry, regulators, and government, the Bank’s strategy aligns with the UK’s ambition to become the world’s most technologically advanced financial centre. Read more here.
Bangladesh moves towards digital financial reporting with Inline XBRL. At a high-level meeting last week, the Bangladesh Securities and Exchange Commission (BSEC) brought together regulators, auditors, exchanges, and corporate leaders to chart a path towards structured, machine-readable financial reporting. There was strong consensus on the need to enhance regulatory oversight and investor confidence through greater transparency. A phased rollout is planned, starting with dual PDF and iXBRL filings from top listed companies, before extending to mutual funds, banks, insurers, and brokers. To support this, a dedicated Taxonomy Committee will adapt international standards for local use, alongside the development of a national platform to centralise and streamline reporting. Read more here.
Securities and Exchange Board of India (SEBI) preparing action plan for quantum-safe computing: Tuhin Kanta Pandey. SEBI is planning for quantum resilience by 2028–29, with a shift to post-quantum cryptography to protect financial systems from emerging cyber threats. Speaking at the Global Fintech Fest, Chairman Tuhin Kanta Pandey stressed the urgency of updating outdated encryption, likening the challenge to Y2K. He also outlined SEBI’s broader tech strategy, including AI-driven fraud detection, real-time oversight, and enhanced cybersecurity frameworks. Read more here.
Monetary Authority of Singapore (MAS) launches BLOOM (Borderless, Liquid, Open, Online, Multi-currency) initiative to extend settlement capabilities. The MAS announced the new initiative, BLOOM, to extend settlement capabilities offered by financial institutions. Through BLOOM, MAS will collaborate with the financial industry to enable settlement in tokenised bank liabilities and well-regulated stablecoins, whilst effectively managing risks in the rapidly evolving digital settlement asset landscape, via standardised approaches. BLOOM builds upon the groundwork established through Project Orchid, which explored use cases for a digital Singapore dollar and the infrastructure required to support it. Read more here.
European Banking Authority (EBA) report highlights supervisory lessons for tackling money laundering risks in crypto-asset services. The EBA report outlines how crypto-asset firms have exploited regulatory gaps through practices like unauthorised operations and forum shopping. It highlights how the European Union’s Markets in Crypto-Assets (MiCA) regulation and updated anti-money laundering and countering the financing of terrorism (AML/CFT) rules aim to address these risks through harmonised authorisation, greater transparency, and stronger governance. Effective implementation will depend on cross-border coordination, ongoing risk monitoring, and robust supervision. Read more here.
International Monetary Fund (IMF) working paper outlines AI project framework for financial supervisory authorities. The paper highlights the growing need for financial supervisory authorities to adopt AI as financial services become increasingly digitalised. It introduces a tailored project management methodology to guide the safe and strategic implementation of AI. Key challenges, including ensuring explainability and mitigating bias, with a focus on stakeholder collaboration, are emphasised, alongside prerequisites for successful deployment, such as robust governance frameworks and adequate resources. Read more here.
Financial Stability Board (FSB) report examines monitoring adoption of AI and related vulnerabilities in the financial sector. The report explores how authorities across member jurisdictions are approaching the oversight of AI in the financial sector. It outlines key considerations and potential indicators for tracking AI adoption and associated vulnerabilities. It also includes a case study focusing on monitoring AI-related third-party dependencies and service provider concentration. The report provides high-level considerations to enhance monitoring and address data gaps. Read more here.
Bank for International Settlements (BIS) releases findings from Project Symbiosis. This joint initiative by the BIS Innovation Hub Hong Kong Centre and the Hong Kong Monetary Authority (HKMA) uses AI and big data to enhance supply chain sustainability disclosures. Project Symbiosis showed that AI and related technology approaches can help improve the speed and quality of environmental data collection, and generate intelligent environmental impact results that are flexible with respect to a wide range of data availability scenarios. Its findings also highlighted financeable emission reduction opportunities. Read more here.
AMLNet: A knowledge-based multi-agent framework to generate and detect realistic money laundering transactions. This paper presents AMLNet, a knowledge-based multi-agent framework designed to address the lack of publicly shareable, regulation-aligned datasets in anti-money laundering (AML) research. The system features a regulation-aware transaction generator producing over one million synthetic transactions across all core laundering stages and typologies, alongside a detection ensemble achieving an F1 score of 0.90. With 75% alignment to AUSTRAC rules and a technical fidelity score of 0.75, AMLNet balances regulatory relevance with behavioural and structural realism. Read more here.
Suptech vendor IRIS’ inaugural newsletter issue explores AI’s expanding role in suptech. The first edition, themed “AI in SupTech: Buzzword or Backbone?”, delves into the shift from experimental AI use to practical implementation in regulatory oversight. It highlights how technologies like AI and XBRL are enabling smarter reporting, market monitoring, and analysis of unstructured data, while also addressing risks such as algorithmic bias and data privacy. With global regulators at varying stages of adoption, ‘The IRIS Lens’ offers insights into both the promise and the pitfalls of AI-powered supervision. Read more here.
MAS highlights industry collaboration and launches PathFin.ai Knowledge Hub at Institute of Banking and Finance Distinction (IBF) Distinction Evening. At the IBF Distinction Evening, the MAS Deputy Chairman stressed the importance of collaboration, continuous skills upgrading, and ethical responsibility in navigating advances in AI. He introduced PathFin.ai, a new knowledge hub aimed at facilitating information sharing on AI implementation experiences and accelerating innovation across the finance industry. The initiative reflects MAS’s commitment to building a resilient, future-ready financial ecosystem through technology and cooperation. Read more here.
Financial Conduct Authority (FCA) announces partnership to accelerate delivery of open finance. The FCA has unveiled a key partnership and launched two techsprints to advance open finance, building on the foundations of open banking. As part of its wider digital innovation agenda, the FCA’s new Smart Data Accelerator will test real-world applications of open finance, supported by a collaboration with data-sharing platform Raidiam. With upcoming sprints focused on mortgages, and small and medium-sized enterprise finance, the regulator aims to foster secure, consumer-focused solutions while preparing to publish a full open finance roadmap by March 2026. Read more here.
AI challenges supervisors’ grip on financial stability, Forum on Financial Supervision article warns. The rapid adoption of AI across financial institutions is outpacing supervisory capacity, threatening systemic stability. AI’s speed, complexity, and opacity give firms a decisive advantage over traditional, human-led supervision. To close this gap, authorities must embed AI directly into supervisory functions, develop secure domestic AI systems, and use advanced techniques like federated learning and real-time AI-to-AI interfaces. Without urgent reform, the current supervisory structure may prove too slow and outdated to prevent future crises. Read more here.
BIS report highlights growing role of AI in central banking and financial oversight. The report, presented to G20 finance ministers and Central Bank governors, details how central banks and regulators are using AI, including generative models, to improve data analysis and policy decisions. While showcasing real-world applications, it also notes challenges such as data governance, skills gaps, and infrastructure needs. To mitigate trade-offs and overcome challenges, collaboration and the sharing of experiences emerge as key avenues. Read more here.
Hong Kong Monetary Authority (HKMA) announces second cohort of generative AI sandbox to advance responsible A.I. innovation. Building on the success of its initial phase, the second cohort of Hong Kong’s genAI sandbox signals a move towards practical and secure A.I. integration in finance. With 27 use cases selected from over 60 proposals, the initiative features strong industry participation and a clear focus on governance, including innovative “A.I. vs. A.I.” quality checks and defences against deepfake-related fraud. Trials are set to begin in early 2026 via Cyberport’s Artificial Intelligence Supercomputing Centre. Read more here.
Bangladesh Bank to launch an interoperable payment system with Gates Foundation support. Speaking at a stakeholder discussion on 15 September 2025, Governor Ahsan H Mansur outlined a renewed initiative to enable instant transfers from mobile wallets, bank accounts, non-bank accounts, or institutional accounts directly to any other account. Backed by the Gates Foundation through the Mojaloop Foundation, the project aims to overcome past implementation challenges and create a more inclusive, accessible financial ecosystem for Bangladesh. Read more here.
Reserve Bank of India (RBI) launches digital payment innovations, integrating AI and internet of things (IoT) with Unified Payments Interface (UPI). The RBI has unveiled four digital payment initiatives at the Global Fintech Fest to enhance customer convenience. These include AI-based UPI HELP for payment assistance and dispute resolution, and IoT Payments for transactions from connected devices. Additionally, ‘Banking Connect’ offers interoperable net banking, and ‘UPI Reserve Pay’ allows users to manage credit limits for specific purposes. Read more here.
France and Slovenia join European Central Bank’s (ECB’s) DLT settlement project. France and Slovenia’s treasuries have joined the Eurosystem’s latest initiative to explore DLT for settling wholesale financial transactions in euro. They are the first central government bodies to participate directly in the ECB-led ‘Pontes’ project, which aims to integrate central bank money with market DLT platforms. The project’s market contact group also includes national central banks from ten eurozone countries, the European Investment Bank, and private sector participants, reflecting a broad effort to modernise Europe’s financial infrastructure. Read more here.
Events
AI Week 2025 opens registration with 200+ free talks on AI innovation. Registration is now open for AI Week 2025, the UK’s largest online festival of AI learning, running from 13–24 October. Hosted by Digital Leaders, the 10-day event will feature 240 expert speakers covering key themes such as ethics, cyber resilience, productivity, regulation, and workforce skills. A standout session led by the Cambridge SupTech Lab and Digital Transformation Solutions will explore AI readiness in financial supervision, drawing on insights from the 2024 State of SupTech Report, which shows a sharp rise in generative AI use and highlights data governance as a major challenge. Read more here.
Central Banks to explore future-proof reserve management in webinar on 29 October 2025. With financial markets becoming increasingly complex and digitised, central banks are reassessing how to modernise reserve and monetary operations. A webinar hosted by Central Banking and sponsored by Nasdaq will examine practical strategies for streamlining operations through cloud-native platforms, AI-driven insights, and modular infrastructure. The session will also address how central banks can prepare for digital currencies, tokenised assets, and shifting regulatory demands, while maintaining control, stability, and transparency. Key insights will include lessons from early adopters and agile approaches to transformation. Read more here.
Suptech is now core to modern supervision, notes Swiss Financial Market Supervisory Authority (FINMA) Chair. At the Global Fintech Fest 2025 in Mumbai, FINMA Chair Marlene Amstad presented findings from the International Organization of Securities Commissions’ (IOSCO) first global survey on suptech, led by FINMA. Covering jurisdictions representing over 75% of global securities markets, the survey reveals that suptech is increasingly integrated into core supervisory work, driven by AI, better data access, and cloud infrastructure, though cyber risks and third-party dependencies remain key concerns. Amstad also highlighted how smaller regulators are using modular solutions to manage limited resources, and outlined FINMA’s own initiatives, including machine learning for market surveillance and open-source tools for insurance stress testing, supported by secure data platforms. Read more here.
