Canonical to the main page Bank of the Future Page

Citi GPS: Global Perspectives & Solutions March 2018 © 2018 Citigroup 26 AI-driven Applications in Banking AI is increasingly seen as a competitive advantage in finance, with banks seeking to analyze structured/unstructured data, turning raw data into actionable intelligence to improve revenues, reduce losses and costs, or do all of the above. We believe most AI implementations in finance are currently focused in these areas:  Improving customer experience – AI applications are helping financial institutions increase customer engagement by providing data insights into user behavior and spending habits, enabling financial institutions to offer contextual, custom recommendations.  AI-based fraud detection – Banks are using AI to analyze client and employee behavior by extracting patterns from huge amounts of unorganized data, which can help identify potential fraud and mitigate risks. AI in cards also helps reduce false declines and increase accuracy of real-time approvals for genuine transactions.  Meeting regulatory requirements, compliance – AI technologies can help financial institutions monitor real-time data and meet regulatory requirements, as well as reduce risks associated with human error and misconduct.  Making data-driven decisions faster – AI can help firms increase human productivity by reducing the amount of time spent on manual/repetitive tasks, resulting in accelerated decision insights. For example, AI-driven credit-scoring methodologies and advisory services augmented with robo-advisors.  Cost savings through automation – AI can help banks streamline their front end by deploying machine learning/robotics to do routine functions and resolve client queries (such as chat-bots), thus helping free valuable human resources for more value-added customer services and improve cost efficiencies. Figure 18. Artificial Intelligence to Banking Source: Citi Digital Strategy, Citi Research Enhance efficiency Scale non-linearly Accelerate & Enhance decision insights Customer Engagement Operations Risk & Compliance • Increase speed & quality of insights to target & personalize • Recommender systems increase targeting returns by 2X+ • Automate & standardize process flows • RPA reduces processing costs by 40-50%; • Reduce time to detection and mitigation • DL techniques can reduce time to fraud detection by 30%+ • Reduce costs by automating manual tracking & reviews • Automation of standard legal processing can reduce costs by 80%+ • Augment and enhance human effort in interactions • Virtual agents can scale to handle 1.5-2MM+ customer queries daily • Augment capacity to address variable demand & complexity • Quicker ledger reconciliations can reduce capacity needs • Simplify and automate user engagement Robo-advisory is 80%+ more cost effective than traditional models • Track risks in real time, at scale; while improving effectiveness • AI use in cyber-security can improve detection by 3x, and reduce false positives • Reduce time to predictions & information retrievals • ML can improve treasury liquidity forecasting effectiveness by 15%+ Simplify; Automate; Efficiency; of existing operating models Meet demand growth without equivalent supply investments VALUE FROM AI TO BANKING Improve decision making; in real time, in complex areas

RkJQdWJsaXNoZXIy MjE5MzU5