By Kevin Telford1
1Advisor, Thoughtworks
Technology continues to grow exponentially into all aspects of our life. Increasingly, financial decisions are made by machines harnessing data and honed by human construct. An ethics void thrives in an invisible, untested, dark data world of science, where artificial intelligence (AI) and machine learning (ML) reside. Investors, entrepreneurs and innovators are increasingly curious about the value of a principled digital application of trust, transparency, ethics on AL and ML. Visionaries see the value for people and the increased returns trusted organizations will deliver. This chapter, titled beyond AI Reg-Tech takes us to a vision where trusted financial services (FS) organizations embrace a standard of trust, ethics, transparency and enablement (TETE) framework. Leading organizations to an AI/ML “kitemark” standard.
For investors, entrepreneurs and visionaries, change has many agents, including legislation such as the EU second Payment Services Directive (PSD2) and the General Data Protection Regulation (GDPR) and the spreading of the Internet of Things (IoT), open banking and connected life – all this maximizes opportunities for innovation and growth. At the intersection of creativity and enabling technology lies an abundance of new products, services and infinite new data sources. We label this “intelligent empowerment” (IE) human and tech augmentation. AI/ML are “components” within IE. For many organizations’ old models, processes are replaced by AI/ML, but this often comes with trade-offs and risk. IE brings mass data integration opportunities from multiple points and combining existing data and enabling FS innovators to bring forward use cases for AI/ML, creating new service efficiencies and products. At the heart is a determination to protect the citizen. Combined data points bring a new age for intelligent analytics, smarter data, faster outcomes that has the investor, innovator community in a frenzy. Money is flowing into new data, consent, privacy protection and the ability to construct valuable new products and services to solve customer challenges.
IE brings complexity to FS and trust. Open banking and IoT increasing connections and integration with FS services through AL/ML has the potential to damage trust irreparably.
Arguably responsible FS organizations build the rationale for trust, transparency and ethics into their models now. Consumer legislation sees to that. The flipside to organizational trustworthiness is that legislation has not kept pace with IE/AI/ML growth.
Accessing code is beyond most people. In any event there is no set of rules to publish the contents in an understandable format. Some organizations cite a trade-off between consumer benefits and privacy. Lazy responses highlight an attitude to trust that is difficult to understand; however, outdated business models die. There is no cop out on trust, transparency and ethics. Some people ask whether an organization can be trusted, and that is what some FS incumbents find, as it is difficult to pivot towards trust as a service. Many FS organizations fail to serve, neglect the risky, underserved and underbanked. There is one big issue: what is the social norm for AI/ML? The challenge lies in human construct and AI/ML intent. Measuring trust will pivot towards the citizen judging FS AI/ML organizations as trustworthy or not.
The proposal is an AI/ML trust framework to encompass the growth of IE, for measuring organizational trustworthiness. The solution is a trustworthy AI framework, designed, adopted, promoted and accepted. For investors, entrepreneurs, visionaries the TETE framework opens new business models. TETE will cover aspects of:
The detail behind TETE will start with human questions. Why are we doing this? Is it ethical? What are the benefits and is it TETE-measurable?
TETE exposes the computational construct:
TETE brings visibility in such areas as
The results will inform a model for TETE. A model that can be used to assess where an organization fits as being AI/ML, IE-compliant and responsible. People will be able to measure a trustworthy entity. For other parties, it is the ability to hold to account organizations deploying AI/ML science so that trust, transparency and ethics become a standard of measurement.
The starting points for an ethical IE/AI/ML framework already exist within international legislation and principles. The future of IE/AI/ML regulation is much publicized. It is largely accepted that a version of a “TETE” framework is inevitable as governments, organizations and the public are increasingly recognizing the issues IE/AI/ML can bring. As AI tech and IoT deployment carry on at pace, events such as social profiling of citizens’ data for FS decisions highlights FS challenges including:
As automated decisions impact citizens’ access to basic rights such as fair finances, then aspects such as creditworthiness have a huge impact on outcomes. We need TETE applied where data is stored, shared or used to construct and conduct automated decisions. For FS, the invisible data decisions applied through IE/AI/ML move further from the “reassurance” of human interfaces for some of the most important decisions humans make. Many FS organizations are faced with the trust legacy, the growth of greenfield entrants starting with the same perspective on trust. It feels to those organizations tangible and measurable. It is, but only one way.
TETE will be implemented as a global standard for FS. It will start with FS self-governing, using the TETE principles through organization collaboration, such as FCA and ICO.
The end solution will be akin to the food label standards and governed under the same principles as domain name management. The aim is for a kitemark, traffic light system and iconic descriptors within interactions, for example, during application.
There is a swathe of entrepreneurial and global government innovation to assist in implementing TETE for the growth of IE. Society is ready; they are awake to privacy issues and increasingly applying human trust measures on their own values and the communities in which they participate. All parties are ready to collaborate around the immense opportunity AI/ML brings; addressing the risk is part of the solution. TETE centres of excellence are a natural next step solution. Shared sandboxes of anonymized human data and open source software. An environment for FS to share data on the same principles as credit reference agencies. Only the purpose is for retrospective and future applications of IE/AI/ML to design, experiment and prove ahead of market launch.
In conclusion, the TETE blueprint is one of many disparate attempts to bring FS self-regulation and the creation of a framework that supports the growth of IE. The solution is centred on protecting the citizen, providing a consistent framework and a level playing field for FS. During research it is almost universally acknowledged that legislation will be required for the growth of IE and AI/ML as components. TETE is an invitation for FS to be proactive and take advantage of the movement for trusted organizations by society. Next steps collaborate and test the principle for the future design of a TETE framework. The TETE framework will bring tangible benefits for all parties, sound business sense and a differentiator as ethics is mainstream. A bigger issue is what is trust and ethics, as the definition knows no boundaries for what is socially acceptable for one or many. The TETE future will enable people to recognize organizations built on trust and decide where to place their business.
Growing the artificial intelligence industry in the UK, Professor Dame Wendy Hall and Jérôme Pesenti:
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/652097/Growing_the_artificial_intelligence_industry_in_the_UK.pdf; https://ai-hr.cyber.harvard.edu/.
AI in the UK: ready, willing and able? https://publications.parliament.uk/pa/ld201719/ldselect/ldai/100/100.pdf.
Government Artificial Intelligence Readiness Index 2019 – Compiled by Oxford Insights and the International Development Research Centre: https://www.oxfordinsights.com/ai-readiness2019 Richard Stirling, Hannah Miller and Emma Martinho-Truswell.
PwC’s Responsible AI: AI you can trust: https://www.pwc.com/gx/en/issues/data-and-analytics/artificial-intelligence/what-is-responsible-ai.html.
Exploring Ethics: An Introductory Anthology by Steven M. Cahn, Oxford University Press – three key levels: applied, normative, and metaethics. Our goal is to understand the significance of ethical thinking in our daily lives.
Artificial Intelligence Benchmark Cap Gemini: www.capgemini.com/wp-content/uploads/2018/07/AI-Readiness-Benchmark-POV.pdf.
UK Can Lead the Way on AI:
Centre for Data Ethics and Innovation (CDEI):
UK AI industrial Strategy www.gov.uk/government/groups/centre-for-data-ethics-and-innovation-cdei.