Edition in partnership with

Welcome executives and professionals. Many studies of AI and work begin with a fixed list of tasks and ask which ones a model can perform. Evidence suggests that the list itself is changing.

Since the previous edition, we have reviewed hundreds of the latest insights in agentic and generative AI, spanning best practices, case studies, market dynamics, and innovations.

This briefing outlines what is driving material value — and why it’s important.

In today’s briefing:

• OpenAI: How AI is expanding work.

• The AI-powered transformation office.

• How to pick your AI sales strategy.

• IBM: Redesign for enterprise AI.

• Transformation and technology in the news.

• Insights for Executive+ members.

• Career opportunities & events.

Read time: 4 minutes.

MARKET INSIGHT

OpenAI: How AI is expanding work

Brief: OpenAI published a 16-page report, drawing on a sample of 800,000 US work-related ChatGPT messages, showing how workers routinely reach beyond the traditional boundaries of their roles, a pattern it calls task crossover.

Breakdown:

• 16.8% of all work-related messages, and 43.5% of occupation-specific ones, concern tasks historically tied to another occupation.

• Patterns of AI use differ across occupations, with marketing and engineering tasks travelling to many workers in other occupations.

• Workers in design, sales and human resources are the most likely to take on a wide variety of tasks drawn from other occupations.

• The effect is stronger at smaller firms, where typical users may turn to AI to expand the work they do when resources are scarcer.

Why it’s important: Usage data like this indicates where work is shifting. It reveals how AI lets workers test new combinations of tasks before firms rewrite job descriptions and roles, an early signal of occupational change that conventional labor-market statistics won't capture until later.

IN PARTNERSHIP WITH RESOLVE

IT automation that pays for itself

Brief: Half your ticket volume is the same problems solved the same way, quietly draining budget. Resolve's agentic automation and orchestration maps every failure path and executes the fix before a ticket ever opens.

The Resolve Platform edge:

• Repeat incidents resolved autonomously before a ticket is ever created

• Failure paths mapped and orchestrated across the entire IT estate

• Engineering hours and budget returned to work that grows the business

Calculate your ROI → Zero Touch Resolution Calculator.

BEST PRACTICE INSIGHT

The AI-powered transformation office

Brief: BCG detailed how transformation offices, which vary by organization but typically cover three core areas: program management, financial and impact tracking, and change management, will be reshaped by agentic AI by 2030.

Breakdown:

• By 2030 the office will apply AI to governance, updates, issue detection and sentiment analysis, freeing staff for more strategic roles.

• Once characterized by manual tracking and reactive reporting, program management will provide always-on issue detection led by AI.

• Financial tracking shifts from periodic validation to ongoing transparency on value, with dynamic forecasting and rapid correction.

• Change management abandons broad, generic interventions for precision engagement, using real-time sentiment and targeted nudges.

Why it’s important: Transformation teams can partner with workstream leaders to design targeted interventions where the value is greatest, rather than monitoring performance and flagging issues for others to resolve. CEOs and the C-suite should be plotting the steps now.

BEST PRACTICE INSIGHT

How to pick your AI sales strategy

Brief: Andreessen Horowitz explored two questions that help decide which AI sales strategy to pick when selling to enterprises: how exposed is the buyer who signs, and does social proof travel?

Breakdown:

• When exposure is high and proof travels, you are in lighthouse territory: a few credible buyers committing first unlock the whole market.

• When mistakes are recoverable and proof travels less, this is landgrab territory: math closes the deal and coverage wins the market.

• When proof travels but isn't required, you may not initially need a large sales team: it spreads engineer to engineer, bottom-up.

• When buyers demand proof but your customer logos don't influence future buyers, you're in a hard market.

Why it's important: Enterprise sales is about risk and reward. Behind every deal is a person who signs their name and wants what they approved to work, and to still have a job next year. They are weighing personal exposure and what evidence would make it bearable.

AI-NATIVE PROFESSIONAL

How to size a market with Claude Cowork

Brief: In this guide, you'll learn to use Claude Cowork for end-to-end market sizing across research, analysis and deliverables. Start with existing company data or research, or let it work from scratch.

Step-by-step:

1. Describe the market you're sizing and define outputs such as a presentation, an Excel workbook, a source document with citations.

2. After your initial prompt, Claude may ask you questions, like market focus and geographic scope, then builds a plan in the sidebar.

3. Claude researches the market, applies the TAM/SAM/SOM framework, then creates three aligned deliverables.

4. Continue the conversation to drill into a specific segment, build a sensitivity model, or generate competitor profiles.

Best practice: Start another task while the analysis runs, review the plan before launch, and cross-check bottom-up calculations against analyst reports.

For the full guide, including board meeting prompts, upgrade to Executive+ or The Boardroom.

MARKET & BEST PRACTICE INSIGHT

IBM: Redesign for enterprise AI

Brief: IBM and Oxford Economics surveyed 1,000 technology CxOs, AI transformation officers and business sponsors on their progress with enterprise AI, highlighting the need to focus on critical value streams to drive growth.

Breakdown:

• At the end of 2025, only 37% of AI initiatives had delivered the business outcomes that senior leadership had originally expected.

• In 2025, 21% of AI ROI was lost to friction across IT and the business, with the failure to redesign processes the top obstacle.

• Improving AI outcomes requires IT upgrades, yet 58% of organizations are under pressure to cut non-AI enterprise tech spending.

• Only 17% of executives say AI approval and funding is consistent across the board, and just 25% measure business value consistently.

Why it’s important: AI agents do their best work at scale, yet fragmented ownership, disconnected systems and broken handoffs are persistent barriers. The greatest returns will go to organizations that redesign value streams to cut friction and let agents move work from one stage of a process to the next.

Futurum published an 8-page case study on IBM's Client Zero program, which captured $4.5B in annualized savings over three years.

Salesforce published a 12-page CX playbook with four plays and a 90-day plan to cut handling time and lift first-contact resolution.

Bain argued governance belongs in the platform control plane, since policy documents apply only when someone remembers to check.

EY set out six dimensions for measuring agenticness and warned AI security is now a board-level risk needing controls by design.

Stanford HAI published a 16-page brief on world models, warning no benchmark lets policymakers judge them for safety-critical use.

Artefact argued you can build a working agent in an afternoon, but production hardening takes 10 to 20 times the prototype effort.

Moonshot AI released the weights and 47-page tech report behind Kimi K3, the largest open-weight model ever, at 2.8T parameters.

Anthropic CEO Dario Amodei published a post setting out the company's open-weights position as it holds out on the industry letter.

Google DeepMind revealed its open-source Gemma model series has passed 900M downloads, with Gemma 4 alone accounting for over 300M.

Meta CEO Mark Zuckerberg argued in a WSJ op-ed that open access, not control by a few firms, is what keeps superintelligence safe. 

Microsoft explored how GeoAI expands enterprise AI, and introduced Project Perception, a continuously learning system of defense.

Palo Alto Networks launched Prisma AIRS AI Gateway, a control plane for enterprise AI, six weeks after closing its Portkey deal.

CAREER OPPORTUNITIES

BlackRock - Head of AI Transformation

JPMorganChase - AI Transformation Executive Director

Google - Enterprise AI Director

EVENTS

Salesforce - Dreamforce - September 15-17, 2026

AWS - Agentic AI on CPU - September 22, 2026

Gartner IT Symposium/Xpo™ - 19-22 October, 2026

Reach enterprise AI decision-makers:

• 66% of readers are C-level executives or VP and Director-level leaders.

• 63.2% of the audience is based in the U.S., EU, UK, ANZ, and Singapore.

• Read by leaders at Microsoft, Deloitte, the Fortune 500, and more.

Guaranteed impression and custom sponsorship packages available, with post-send performance reporting.

Conceived as a practical communication for executives Lewis Walker has worked with, this briefing has become a trusted resource for thousands of senior decision-makers shaping the future of enterprise AI.

We welcome your feedback.