0727

AI Insights & Research

Bank podcast transcript | Scenario analysisMorgan Stanley·

What would make AI infrastructure investment pay off?

Brian Nowak compares compute rental, model services on owned infrastructure and model services using rented capacity. Pricing power and processing efficiency drive the resulting economics. The returns discussed are scenario estimates rather than realized results; the scale of capital expenditure alone cannot establish investment success.

Source
Analyst perspective | Cybersecurity market analysisMorgan Stanley·

Agent identity becomes an enterprise security problem

Morgan Stanley argues that agents require identity controls covering purpose, duration and permitted actions, with decisions made during execution across systems. This could expand identity-security demand. Market-size estimates are analyst forecasts, not realized demand or evidence that a particular security product prevents agent misuse.

Source
Bank research commentary | Cross-country observational analysisGoldman Sachs Research·

AI employment pressures remain concentrated, Goldman Sachs finds

Goldman Sachs combines adoption surveys with occupational employment data and finds weaker hiring in more exposed sectors, with potentially greater pressure on junior workers. Economy-wide effects remain limited in its estimates. These associations and model-based comparisons do not attribute every departure from employment trends to AI.

Source
Provider threat-intelligence case reportAnthropic·

From answering questions to executing workflows: Monitoring AI misuse

Anthropic reviews misuse it detected and disrupted between December 2025 and August 2026, including workflow execution, fraud and surveillance. The cases motivate ongoing monitoring. They are selected observations within one provider’s visibility, not prevalence estimates or evidence that humans have left the command chain.

Source
Enterprise survey and consulting analysisBain & Company·

Bain: Why rising AI budgets can miss their return targets

Bain’s survey of 951 companies identifies a gap between savings targets and reported outcomes, with many agents still requiring human approval or exception handling. The authors recommend reflecting oversight, data integration and workflow redesign in investment cases. Survey associations and consulting experience do not establish a single cause of success or failure.

Source
Industry outlook and scenario estimateBain & Company·

AI may reshape where profits accrue, not just production costs

Bain separates AI’s industry impact into productivity, innovation and shifts in competitive share. Its $4.7 trillion estimate concerns profit pools over 2025–2035 and includes creation and redistribution of profits. It is a forward-looking analytical estimate, not realized net gains or a certain forecast.

Source
Small executive survey and perspectiveBoston Consulting Group·

Delegating to AI while preserving opportunities to practice

About half of 70 executives surveyed by BCG reported observing deskilling. The authors argue that judgment, problem framing and creativity need continued practice within everyday work. These findings reflect a small executive sample’s perceptions, not direct measurement of workers’ cognitive decline.

Source
Executive survey reportBoston Consulting Group·

BCG: AI investment puts organizational responsibility on CEOs

BCG’s AI Radar surveyed 2,360 executives, including 640 CEOs. Nearly three-quarters of CEOs described themselves as their organization’s main AI decision maker. The report frames AI as a strategy, workforce and governance issue. Investment intentions and confidence in returns are expectations, not realized financial results.

Source
Enterprise governance perspectiveBoston Consulting Group·

What should IT govern when business users build their own agents?

BCG argues that business-built CRM agents introduce risks around customer records, consequential decisions and consumption costs. It recommends shared data architecture, agent governance and cost controls while moving development closer to business users. This is practice-based governance guidance, not a universally validated deployment template.

Source
Consulting practice perspectiveBoston Consulting Group·

AI in the boardroom: Bring evidence and assumptions into the room

BCG describes decision agents that assemble cross-functional evidence, identify missing data and compare scenarios under different assumptions. The proposed value is better-informed human deliberation, not automatic transfer of strategic accountability. This consulting perspective does not establish a general causal improvement in boardroom decision quality.

Source
Stay curious. See you in the next read.

Join the conversation after approval.

Comments, replies and messages are reserved for approved, signed-in collaborators.

  1. Submit your collaborator profile
  2. Wait for profile review
  3. Sign in after approval to participate
Already a member? Sign inBecome a collaboratorThis is a flow preview. Account sign-in and approval checks are not connected yet.