MAS - Monetary Authority of Singapore
MAS - Monetary Authority of Singapore
11/09/2021 | Press release | Distributed by Public on 11/09/2021 01:53
MAS Announces Winners of the Global Veritas Challenge 2021 at Singapore FinTech Festival
Singapore, 9 November 2021… The Monetary Authority of Singapore (MAS) today announced the three winners of the inaugural Global VeritasVeritas is a framework which enables financial institutions to evaluate their AIDA-driven solutions against the FEAT principles. The Veritas framework was highlighted by Mr Heng Swee Keat, Deputy Prime Minister of Singapore, in his speech laying out Singapore's National AI Strategy at SFF x SWITCH in 2019. Challenge 2021. The winners, selected from a pool of ten finalists, successfully addressed key challenges in validating the fairness of artificial intelligence (AI) solutions for specific banking use cases.
2 This year's Challenge focused on four use cases, namely (i) product marketing; (ii) risk, compliance and fraud monitoring; (iii) loan origination and know-your-customer (KYC); and (iv) credit scoring and profiling. Validating the fairness of artificial intelligence and data analytics (AIDA) solutions in these areas ensures that AIDA-driven decisions do not result in particular groups of customers being systemically disadvantaged. This fosters trust in the solutions and promotes greater development and adoption of AI, allowing financial institutions to harness the technology's benefits more broadly while guarding against its risks.
3 The winners were selected during the Global Veritas Challenge Demo DayPlease access the Global Veritas Challenge Demo Day video here. held on 20 October 2021, where finalists pitched their innovations to a judging panelPlease see Annex B for the full list of judges.. The finalists came from six countries in Asia, North America and Europe and were shortlisted from 74 submissions. Each winner received a cash prize of S$50,000. The top three teams are (in no order of merit):
2 This year's Challenge focused on four use cases, namely (i) product marketing; (ii) risk, compliance and fraud monitoring; (iii) loan origination and know-your-customer (KYC); and (iv) credit scoring and profiling. Validating the fairness of artificial intelligence and data analytics (AIDA) solutions in these areas ensures that AIDA-driven decisions do not result in particular groups of customers being systemically disadvantaged. This fosters trust in the solutions and promotes greater development and adoption of AI, allowing financial institutions to harness the technology's benefits more broadly while guarding against its risks.
3 The winners were selected during the Global Veritas Challenge Demo DayPlease access the Global Veritas Challenge Demo Day video here. held on 20 October 2021, where finalists pitched their innovations to a judging panelPlease see Annex B for the full list of judges.. The finalists came from six countries in Asia, North America and Europe and were shortlisted from 74 submissions. Each winner received a cash prize of S$50,000. The top three teams are (in no order of merit):
| Company Name | Challenge Theme | Project Summary |
| Cylynx Pte Ltd | Risk, Compliance and Fraud Monitoring |
Aim: Integrate fairness into AIDA model development and live production monitoring for fraud detection. Capabilities: Enables business teams to flag out individuals or groups that may be systematically disadvantaged and set thresholds to avoid unintended biases. |
| Visa Inc. | Product Marketing |
Aim: Provide financial institutions with a set of standardised reusable components for screening and mitigating unwanted biases in marketing propensity models. Capabilities: Equip financial institutions with a toolkit to enhance the fairness of their marketing practices, via (i) screen for sensitive attributes and identify favourable/unfavourable outcomes by group; and (ii) remove systemic and harmful biases. |
| TruEra Inc. & Demyst Data Ltd. | Credit Scoring & Profiling |
Aim: Mitigate the effects of bias in credit scoring due to biases in training data or issues with the AIDA model. Capabilities: The project uses external data (via Demyst platform) to build a machine learning-based credit scoring model. It then uses TruEra's AI solution to (i) define a protected group and set a fairness objective (ii) analyse this model for evidence of bias and accuracy issues, (iii) identify root causes contributing to such bias; and (iv) conduct targeted interventions. |
4 Mr Sopnendu Mohanty, Chief FinTech Officer of MAS, said, "We are very encouraged by the creative solutions of the winners and finalists of this year's Global Veritas Challenge. Responsible AI is essential to fostering trust amongst relevant stakeholders including users of AI solutions and recipients of AI-driven decisions. The solutions allow financial institutions to harness the tremendous benefits of AI in their operations while guarding against its risks. We look forward to seeing more innovative solutions in future editions of the Challenge. MAS would also like to thank Accenture for their support in organising this year's Challenge."
5 Mr Lee Joon Seong, Managing Director, Accenture's Applied Intelligence Lead for Southeast Asia, said "As more companies begin to adopt AI in their operations, it is important that we strengthen the trustworthiness of AI and data analytics-driven solutions. This is even more urgent in financial services, where trust is fundamental to banks and financial institutions in their customer transactions. It is encouraging to see many innovations from this year's participants, as we work towards building a financial system that is fair, ethical, accountable and transparent. Accenture looks forward to continuing our partnership with MAS and the wider ecosystem in establishing Singapore as a responsible AI hub for the financial services sector."
***