Key Questions Banks should ask Before Adopting AI/ML

~ By  Chida Sadayappan & Ishita Vyas

Either due to changing customer needs, fluctuating regulatory and compliances, peer pressure or to remain relevant amidst the unprecedented pandemic times, banks have begun to realize the importance of adoption of AI/ML. Whether banks are just starting their AI/ML journey (beginners) or they have come far in their journey, it is essential for them to keep in check if they are “asking the right questions”:

• Are we ready to adopt AI/ML? 
• What to do: Buy vs Build vs Buy then Build?
• How to establish the right culture and proper governance to adopt AI/ML?

 Are we ready to adopt AI/ML?

Before taking the first step towards implementing AI technologies at a larger scale, there are 2 things Banks need to consider:

• Unlocking the value from non-digitally available data:

Since data is the foundational layer upon which information and insights can be generated, it is essential to extract value from the data gathered from non-digital assets, human-based or document-oriented processes also. To reach there, banks need to begin with digitizing processes and documents and automate existing redundant manual processes.

• Revamping the Legacy systems with Data focused strategies:

Many of the banking institutions have been around for more than 2 decades due to which they have a weak portfolio of tightly coupled legacy systems, this has been a constant obstacle for banking institutions to deploy newer technologies. It's unrealistic to replace these legacy systems in a big swoop, instead, banks should implement a phased approach. Starting with better data management and governance, rationalizing IT architecture, adopting cloud for gaining better computing power to enable AI/ML, and embracing DevOps and DevSecOps to automate developments, assuring continuous delivery, innovation and monitoring. 

Transforming old processes and modernizing legacy technologies is easier said than done. However, this is a crucial step in the transformation journey.

What to do: Build vs Buy?

Well, now that we have overcome the initial obstacles, it's time to answer the million-dollar question: “Build vs Buy”. 

Banks need to set an ultimate goal for AI/ML. If banks anticipate becoming AI-first organizations or want to transform low-hanging, cherry-picked use cases that will generate instant value? 

• Buy: AI/ML technology acts as a catalyst to execute routine/repetitive back office or middle office tasks such as processing, reconciliation, AML, etc. more effectively and efficiently, but it’s not recommended to DIY AI model for such use cases. These use cases will deliver quick values, but the time invested in hiring/training the right talent, investment in supporting infrastructure, etc. would be huge as compared to buying an off-the-shelf solution. Even if you choose to build your own AI with open source tools, it can cost millions of dollars and can take months to train machine learning algorithms to do what most vendors have already achieved. 

Commercial AI platforms not only allow teams to complete one data project from start to finish but also introduce efficiencies all over. This includes features like cutting time spent in cleaning data, smoothening production issues and avoiding reinventing the wheel when deploying models on a daily basis or building in the documentation and best practices for enabling reproducibility.

PNC Bank has been working with AI vendor Anaconda since they started their AI journey, to overhaul its data science infrastructure for Python and R. Anaconda claims PNC is currently able to build machine learning models in-house leveraging Anaconda’s open-source platform for use cases like predicting losses, protect the bank, and set prices, and the Management Information Systems group also began making applications in Python that help with banking operations 

• Build: Even though buying an off-the-shelf solution or leveraging open source tools can be quick and convenient but there are times when banks have no choice but to build the model from scratch. A few of the drivers are:
• Use case is so unique that no commercial AI product or tool is available in the market
• Gaining full ownership of the code and model is of utmost importance as per dictated regulations and compliance
• Bank is going for a long-term commitment and perceives deployment of AI/ML technology as a differentiation, also doesn’t want to share its competitive edge with others via vendors pooled data lake

For e.g. JP Morgan Chase & Co.-the American multinational investment bank and financial services company has built an inhouse AI solution called COiN-a short-form for Contract Intelligence platform that uses unsupervised machine learning, which automated the processing of legal documents, extraction of data, and review of certain types of legal contracts—reducing 360,000 person-hours. 

• Buy-then-build: It is also important to identify where do you stand in your AI/ML transformational journey and how mature your use cases are. If you have just commenced on the AI/ML path, with use cases like Automation bots (RPA), AI assistants, computer vision etc., it is better to wet your feet with buying/deploying off the shelf AI tools or leverage open source AI platforms like TensorFlow, Scikit-learn etc. and gradually start building in-house AI/ML capabilities for more unique use cases.

The question is gradually shifting towards 'buy or buy-then-build' as the strategy poses quite a few advantages:

• Smaller initial investment
• Prompt kick start in implementing AI/ML technologies
• Buying time to hire new data scientists or train internal employees
• Learn from the mistakes that occurred during POCs with the vendor
• Better and more data is accessible (considering data pool from the vendor) to use for in-house AI/ML models

How to establish the right culture and proper governance to adopt AI/ML?

As the pace of AI and machine learning adoption among banks is increasing, a lot of limitations have surfaced up, like reluctance from employees, governance of models in operation due to constantly changing regulatory environment, getting models fully integrated into the production system, and ongoing management and maintenance of models. This calls for better governance and a transformed operating model:

• Right culture: Every organization faces resistance and fear from the employees when planning to implement AI and machine learning. This fear is- fear of becoming obsolete. Hence, a major cultural shift is vital. Here the role of leaders of the organization comes into play. They should work to inspire their employees to adopt AI by communicating and demonstrating how AI will enable them to do their jobs better. Educate them to see the benefits of adoption of AI and ML technologies by focusing their use cases on reducing work overload, including repetitive tasks in call centers, or expanding business in a way that adds to employment, for example through new product innovation or by entering new markets. 

For example, a bank created a document for relationship managers that highlighted how combining their expertise and skills with AI’s tailored product recommendations could improve customers’ experiences and increase revenue and profit. The AI implementation plan also included sales competitions based on the use of new tools; the winners’ achievements were showcased in the CEO’s monthly newsletter to employees.

• Proper Governance: As Gartner predicted, by 2023, 60% of organizations with more than 20 data scientists will require a professional code of conduct incorporating ethical use of data and AI. Now that AI adoption is no more an option, leaders need to think about AI governance even before implementing the technology.

Deloitte's Trustworthy AI framework consists six dimensions for organizations to consider when designing, developing, deploying and operating AI systems. The framework helps manage common risks and challenges related to AI ethics and governance, including:

• Fair and impartial use checks: A common problem with AI is to avoid human bias in the coding process. To avoid this, banking institutions need to determine what constitutes fairness and actively identify biases in their algorithms and data and implement control measures to avoid unexpected outcomes.
• Implementing transparency and explainable AI: For AI to be trustworthy, all participants have a right to understand how their data is being used and how the AI system is making decisions. Banks must be prepared to create verifiable algorithms, attributes and correlations.
• Responsibility and accountability: A strong artificial intelligence system must include guidelines that clearly state who is responsible for making these guidelines. This question embodies an unknown aspect of artificial intelligence: are developers, testers or product managers responsible? Do you understand the inner workings? Whose will be ultimately responsible among the C-suite?
• Putting proper security in place: To be reliable, AI needs to be protected from risks, including cybersecurity risks that may cause physical and/or digital harm. Organizations need to carefully check and eliminate all types of risks and communicate these risks to users.
• Monitoring for reliability: For AI to achieve widespread adoption, it must be as robust and reliable as the traditional systems, processes and people it is augmenting. Banks need to ensure their AI algorithms produce the expected results for each new data set. They also need to established processes for any issues and inconsistencies in the expected outcomes
• Safeguarding privacy: Trustworthy AI must comply with data regulations and only use data for its stated and agreed-upon purposes. Banks should ensure that consumer privacy is respected, customer data is not leveraged beyond its intended and stated use, and consumers can opt-in and out of sharing their data.

 To sum it up, irrespective of wherever you are in your AI/ML transformational journey, it's advisable to keep check of if you are asking the right questions and taking care of all the traditional yet foundational pillars: people, process, governance and technology.

Check out newly launched Deloitte’s ReadyAI? Artificial Intelligence-as-a-Service Solution