0727

AI Insights & Research

University policy brief | Legal-text screening researchStanford HAI / RegLab·

AI-assisted review makes outdated local laws visible

Stanford researchers collect laws from 9,623 jurisdictions and use an LLM pipeline to prioritize potentially discriminatory provisions for human verification. The approach may reduce the cost of statutory review. It targets textually identifiable disparate treatment, not all discriminatory effects, and neither replaces legal judgment nor automatically changes the law.

Source
University research report | Algorithm experimentsMIT News·

Generative plans need to satisfy hard constraints

MIT presents HardFlow, which gives a generative model flexibility during sampling while enforcing constraints on its final output and optimizing solution quality. Experiments cover robotic planning, physical-process control and image editing. Constraint satisfaction in these tests is not a blanket safety guarantee for every real deployment environment.

Source
Journal article | Medical-imaging algorithm evaluationNature·

Patient-specific AI speeds up alignment of surgical images

MIT and collaborators introduce xvr, which generates training data from a patient’s preoperative scan and rapidly adapts a model to align intraoperative X-rays with 3D images. Evaluation across anatomical regions and hospitals supports faster, accurate registration. It does not by itself establish fewer surgical complications or improved patient outcomes.

Source
Consulting perspective | Strategy-officer survey and frameworkBoston Consulting Group·

When analysis becomes abundant, strategy needs judgment

Drawing on a survey of 175 chief strategy officers, BCG argues that complexity, reinvention and competing time horizons require more than additional analysis. Strategists should choose suitable modes of thinking and scrutinize AI recommendations. This management framework and interview evidence do not experimentally establish that human judgment always outperforms AI.

Source
Bank industry article | Interviews and practice observationsJ.P. Morgan·

Machine learning helps organize private-credit evidence

J.P. Morgan describes using machine learning to extract, standardize and validate non-standard private-credit documents so analysts can focus on credit judgment. Models may prioritize risks, while experts still interpret outputs and challenge errors. The article draws on industry interviews and experience, not an independent experiment demonstrating improved defaults or returns.

Source
Consulting research | Global workforce surveyBoston Consulting Group·

Time saved by AI needs an organizational plan

BCG surveys nearly 12,000 employees, managers and leaders and finds AI adoption outpacing work redesign. Many report time savings without guidance on reinvestment. Clear strategy, redesigned processes and training are associated with better experiences and value indicators. Self-reported survey relationships do not establish causal returns from a management approach.

Source
Consulting perspective | Implementation experience and illustrative economicsMcKinsey / QuantumBlack·

Measure agent economics by the cost of completed work

McKinsey recommends evaluating the full cost of finishing an onboarding, claim or sale, including human review, exceptions, orchestration and maintenance. Cheaper model calls alone may not lower delivery costs; reuse and workflow redesign also matter. Its examples reflect early experience and illustrative calculations, not guaranteed enterprise returns.

Source
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
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.