Artificial Intelligence enabled solutions will drive productivity in operations in the coming years, along with changes to processes to best utilize these capabilities. Advisory companies have a variety of views on the level of productivity that will be achieved.

Recent studies and reports have provided various estimates of productivity gains from artificial intelligence (AI) in the banking industry, particularly focusing on generative AI (GenAI).

• Bain & Company (2024): A survey of 109 US financial services firms reported an average productivity gain of 20% across various applications, including software development, customer service, and IT, driven by generative AI.
• McKinsey & Company (2023): Estimated that generative AI could deliver $200 billion to $340 billion annually in value for the banking industry if fully implemented, equivalent to 9-15% of operating profits, primarily through productivity improvements in customer operations, risk management, and software development.
• EY Report (2025): Focused on the Indian banking sector, projecting productivity gains of up to 46% in banking operations, 38-40% in sales and customer service, and 34-36% in credit and collections by 2030 due to GenAI adoption.
• Deloitte (2025): Predicted that AI tools could save 20-40% in software investments for the banking industry by 2028, with per-engineer cost savings of $0.5 million to $1.1 million. Specific cases, like Citizens Bank, showed a 20% productivity boost among engineers using generative AI, while industry-wide studies noted 30-55% productivity increases for programmers.
• Accenture (2024): Found that banks implementing generative AI could boost productivity by nearly 30%, with potential revenue increases of up to 6%.
• BCG (2022): Highlighted case studies where banks achieved 10-20% profit increases using advanced analytics, though the impact of AI was noted as marginal compared to its potential.
• Global Finance Magazine (2025): Referenced productivity improvements of 25% for customer care professionals using generative AI for tasks like auto-summarization of calls, with broader estimates aligning with McKinsey’s $200 billion to $340 billion in annual productivity gains.

These estimates vary based on the scope of AI application (e.g., software development, customer service, risk management) and geographic focus. Challenges such as regulatory compliance, data security, and the need for robust governance frameworks can impact the pace and scale of adoption, potentially moderating these gains.

A key headwind to the adoption of AI is general understanding and the wide range of risks associated with use and management. These are some of the most commonly discussed:

Data Privacy and Security

• Risk: AI systems rely on vast datasets, including sensitive customer information (e.g., financial records, personal details). Breaches or misuse could lead to data leaks, violating regulations like GDPR or CCPA.
• Impact: Financial penalties, reputational damage, and loss of customer trust. For example, a 2024 report noted that 60% of banking executives cited data privacy as a top concern for AI deployment.
• Mitigation: Implement robust encryption, anonymize data, and use secure AI platforms with zero-data-retention policies (e.g., Salesforce’s approach).

Model Bias and Fairness

• Risk: AI models trained on biased or incomplete data can produce unfair outcomes, such as discriminatory lending decisions or skewed risk assessments.
• Impact: Regulatory scrutiny, legal challenges, and reputational harm. For instance, biased credit scoring models could violate fair lending laws.
• Mitigation: Use explainable AI, conduct regular bias audits, and ensure diverse datasets, as recommended by vendors like Temenos and IBM.

Regulatory and Compliance Challenges

• Risk: AI applications must comply with complex banking regulations (e.g., AML, KYC, Basel III). Lack of transparency in AI decision-making (“black box” models) can complicate audits.
• Impact: Fines and operational restrictions. A 2025 EY report highlighted that 45% of banks struggle with aligning AI with regulatory requirements.
• Mitigation: Adopt transparent AI solutions (e.g., SymphonyAI’s Sensa Copilot) and integrate compliance-focused tools like those from nCino or Minerva.

Cybersecurity Threats

• Risk: AI systems can be targeted by cyberattacks, such as adversarial attacks that manipulate model outputs or exploit vulnerabilities in AI infrastructure.
• Impact: Compromised fraud detection or unauthorized access to systems, leading to financial losses.
• Mitigation: Deploy AI-driven cybersecurity tools (e.g., Feedzai, Palantir) and conduct regular penetration testing.

Operational and Integration Risks

• Risk: Integrating AI with legacy banking systems can be complex, leading to downtime, errors, or inefficiencies. Overreliance on AI without human oversight can also amplify mistakes.
• Impact: Disrupted operations and increased costs. A 2024 Accenture study noted that 30% of banks faced integration challenges with AI solutions.
• Mitigation: Use modular AI platforms (e.g., Sopra Banking Software) and maintain human-in-the-loop oversight.

Ethical and Reputational Risks

• Risk: Misuse of AI (e.g., overly aggressive customer targeting or lack of transparency) can erode trust. Public perception of AI as replacing human jobs also poses reputational challenges.
• Impact: Customer backlash and loss of market share.
• Mitigation: Adopt ethical AI frameworks, ensure transparency, and communicate AI’s role as an enabler, as seen in Kasisto’s customer-focused chatbots.

Cost and Scalability Issues

• Risk: High upfront costs for AI implementation, coupled with challenges in scaling across diverse banking operations, can strain budgets.
• Impact: Delayed ROI or project abandonment. Deloitte’s 2025 report estimated that 25% of banks faced cost overruns in AI projects.
• Mitigation: Start with pilot projects, leverage cloud-based AI solutions (e.g., Ayasdi), and partner with cost-effective vendors like Appinventiv.

Lack of Skilled Talent

• Risk: Shortage of AI expertise in banking can hinder effective implementation and maintenance of AI systems.
• Impact: Suboptimal AI performance and higher reliance on external vendors.
• Mitigation: Invest in upskilling programs and collaborate with vendors offering managed AI services (e.g., Apexon, IBM).

Broad adoption of AI is accelerating across all industries and banking has historically been a primary user of technology to manage operations and financial risk. We expect this trend to continue and it will be critical for executive teams to continue to expand their understanding of both the risks and benefits.

Source: Grok, Greystone Advisors