BAC emphasizes disciplined, ROI-led AI adoption while containing technology costs through subscription contracts.

Overview

This management-meeting note examines how Bank of America is approaching AI investment, operational efficiency and cybersecurity. Management emphasized disciplined deployment, measurable returns and subscription-based vendor contracts, while Goldman Sachs maintained its Buy rating and $79 12-month price target.

Core views

Bank of America’s technology strategy puts risk, regulatory and security controls ahead of discretionary investment decisions. Management said discretionary spending has increasingly shifted from traditional digital transformation toward AI-focused initiatives, but projects are assessed through an operational-excellence framework rather than pursued simply for adoption. BAC has mapped 3,700 firmwide processes, enabling it to evaluate a proposed project’s workflow impact and whether the expected efficiency gain supports the investment. Management characterized the technology budget as ROI-focused, with some savings from efficiency gains used to self-fund incremental spending. Discretionary technology investment is approximately $4bn within total technology spending of roughly $13bn. Over the past decade, discretionary spending has increased about 45%, compared with about 15% growth in core operational technology spending. For incremental discretionary projects, BAC targets an average payback period of approximately three years. The company described a bottom-up AI deployment model with three levels. Level 1 covers broadly deployable, commoditized employee-productivity tools; Level 2 uses AI features embedded in third-party platforms rather than rebuilding similar capabilities internally; and Level 3 applies custom solutions to complex workflows. Compared with the earlier digitization cycle, which focused mainly on automation and client experience, management expects AI to support more complex internal processes such as compliance and relationship management. The wider potential use case also creates greater implementation complexity because AI is predictive rather than deterministic, requiring deeper process mapping, guardrails and continuous oversight. Management expects AI-related efficiency gains to be meaningful but difficult to quantify directly. Revenue-oriented benefits are harder to measure, while some expense savings are reinvested into further technology investment. It therefore indicated that firm-wide operating metrics—particularly operating leverage and revenue-growth acceleration relative to peers—may offer the more useful evidence of AI-driven productivity. BAC said it has been relatively insulated from the AI-related expense pressure seen elsewhere in the industry from compute demand, token usage and service fees. Its ROI discipline is intended to prevent spending on AI where simpler automation would suffice, and its mainly subscription-based vendor contracts are expected to remain in place over the next few years. This structure gives BAC flexibility to broaden AI use without being constrained by incremental query volume or usage-linked spending. Cybersecurity is the main counterweight to broader AI deployment. Management noted that models capable of improving productivity can also identify and exploit vulnerabilities, increasing both the urgency and complexity of cyber-risk management. BAC is focused on faster vulnerability patching, evaluating risks in aging systems, considering open-source-software exposure and modernizing platforms where necessary. Regulators were described as open-minded but still developing their approach to AI workflow augmentation; they are engaging banks to understand specific use cases and oversight frameworks. Goldman Sachs remains Buy-rated on BAC and sets a 12-month $79 price target based on 15.5x 2027E P/E. Its estimates show revenue rising from $113,706.0mn in 2025 to $123,845.4mn in 2026E, $130,471.0mn in 2027E and $135,891.1mn in 2028E, while EPS is projected at $4.58, $5.10 and $5.64 for 2026E–2028E, respectively. The stated downside risks are less efficiency improvement from NII attrition and slower deposit repricing.

Analysis framework

The note synthesizes management’s discussion of technology strategy, then links project selection, process mapping, payback discipline, vendor contracts and cybersecurity controls to the potential operational effects of AI. Goldman Sachs values BAC using a 2027E P/E multiple and provides forward financial estimates.

Methodology notes

• Valuation methodsP/E and PEG Valuation

  15.5x 2027E P/E valuation

  Goldman Sachs derives its 12-month $79 target price by applying a 15.5x price-to-earnings multiple to its estimated 2027 earnings.

• Corporate Fundamentals and FinanceOperating and Financial Leverage Analysis

  Operating leverage and revenue-growth acceleration versus peers

  Management indicated that firm-wide operating leverage and relative revenue-growth acceleration may be better indicators of AI-related efficiency than directly measuring every individual AI benefit.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Bank of America Corp. (BAC.US)

  Primary covered company; management expects disciplined AI deployment and process re-engineering to support efficiency while subscription contracts contain near-term AI cost exposure.

  Strengths

  Mapped 3,700 processes, ROI-focused project selection, subscription-based AI vendor arrangements and emphasis on cyber controls.

  Weaknesses

  AI efficiency benefits are difficult to quantify directly because revenue benefits are hard to measure and savings may be reinvested.

  Comparison

  Management contrasted AI with the earlier digitization cycle: AI can address more complex workflows but requires more oversight because it is predictive rather than deterministic.

  Risks

  Less efficiency improvement from NII attrition and slower repricing of deposits.

Key data

• Total technology spending~$13bnApproximate total technology spend cited by management.
• Discretionary technology investment~$4bnApproximate portion of total technology spending; has grown ~45% over the past decade.
• Core operational technology growth~15%Growth over the past decade, compared with ~45% for discretionary technology investment.
• Target payback period~3 yearsAverage target for incremental discretionary investments.
• Firmwide processes mapped3,700Used to assess workflow impact and expected returns on proposed projects.
• 2026E / 2027E / 2028E EPS$4.58 / $5.10 / $5.64Goldman Sachs forecasts.
• 2026E / 2027E / 2028E revenue$123,845.4mn / $130,471.0mn / $135,891.1mnGoldman Sachs forecasts; 2025 revenue was $113,706.0mn.
• Price target and valuation basis$79; 15.5x 2027E P/EGoldman Sachs 12-month target; BAC closed at $62.67 on 9 Sep 2026.

Impact & implications

The report presents BAC’s AI strategy as an efficiency opportunity pursued within strict return, risk and governance disciplines. Subscription-based vendor agreements may reduce near-term AI cost pressure, while the scale of gains should be assessed through broader operating performance rather than isolated project metrics.

Risks

• Downside risks include less efficiency improvement from NII attrition and slower repricing of deposits.

What to watch

• Evidence of AI-driven operating leverage and revenue-growth acceleration relative to peers.
• Execution of the three-level AI deployment model and whether projects meet the roughly three-year average payback target.
• The durability of subscription-based AI vendor contracts and their effect on scaling costs.
• Cybersecurity controls, including vulnerability-patching speed, aging-system risk and open-source-software exposure.
• Regulatory development and bank-regulator engagement on AI workflow oversight.