Edition in partnership with

Welcome executives and professionals. The fundamentals of growth still matter, but the way companies create, capture, and scale growth with AI has fundamentally changed the game.

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:

• AI as a growth force multiplier.

• The reality of enterprise AI pricing.

• Can AI agents use computers reliably?

• Governing agentic systems in real time.

• Transformation and technology in the news.

• Insights for Executive+ members.

• Career opportunities & events.

Read time: 4 minutes.

BEST PRACTICE INSIGHT & CASE STUDIES

AI as a growth force multiplier

Brief: McKinsey detailed the myths that hamper growth in the new world of AI, and how overcoming them requires leaders to focus less on technology and more on redesigning how commercial decisions are made.

Breakdown:

• McKinsey sees seven common myths that hamper leaders' ability to turn AI into value (e.g. AI takes a long time to create P&L value).

• Overcoming them requires CEOs and chief commercial officers to compress the distance between customer signal, decision, and action.

• McKinsey recommends that CEOs concentrate their efforts on the one to three domain-level opportunities that carry the most value.

• Turn each opportunity into specific, measurable outcomes on a clear timeline, then mobilize cross-functional teams and resources.

Why it’s important: The companies pulling ahead recognize how deeply AI is changing the way they create demand, convert customers, and generate value. Early AI leaders are achieving EBITDA uplifts of 20 percent or more by rewiring their commercial engines around AI.

BEST PRACTICE INSIGHT

The reality of enterprise AI pricing

Brief: Bain & Company explored the enterprise realities of AI pricing and how capacity, rather than consumption, is emerging as the preferred model, as it offers predictable budgets for customers and stable revenue for vendors.

Breakdown:

• Under effort and output models, vendors are paid whether or not the output creates value. Under an outcome model, payment follows results.

• Outcome-based pricing has found a home in customer support, where a resolved conversation is observable, attributable, and contractible.

• The bigger divide may not be effort vs. output vs. outcomes, but direct usage vs. capacity: four in five AI vendors choose capacity.

• Customers commit to fixed capacity with no rollover or refund, extending seat licensing, where much of what is bought goes unused.

Why it’s important: For enterprises, capacity models create budget predictability: procurement teams and CFOs can approve a defined commitment more easily than open-ended variable spending. For vendors, capacity models preserve many economic advantages that exist under seat-based pricing.

IN PARTNERSHIP WITH THE BOARDROOM

AI blueprints that actually deliver results

Proven results: CXOs, VPs, and directors are leveraging step-by-step AI transformation blueprints not available anywhere else.

• The Q2 2026 blueprint helped a consulting partner increase an AI transformation deal from USD 5.1M to USD 12.6M.

• The Q1 2026 blueprint helped a Fortune 500 CAIO expand AI opportunity pipeline value by 42% and cut projected time-to-production by 28%.

What’s inside The Boardroom:

• Step-by-step executive blueprints for AI transformation

• The Executive AI Index: 320 top 1% AI playbooks

• Full AI-native guides to accelerate your career

• The extended edition of Enterprise AI Executive

The next blueprint will be delivered to members' inboxes on September 21, 2026. If you're looking to drive enterprise AI P&L impact, join us inside The Boardroom.

MARKET & BEST PRACTICE INSIGHT

Can AI agents use computers reliably?

Brief: Andreessen Horowitz explored how computer-using agents are beginning to hold up in production, as the frontier shifts from "can the agent use a computer?" to "can it reliably do this job" inside an actual enterprise?

Breakdown:

• The first wave of computer-use infrastructure made agents capable: seeing, clicking, typing, recovering. The next makes them useful.

• As UI navigation becomes a model-layer commodity, advantage moves up the stack: context, permissions, validation, and error handling.

• A year ago the best model scored 42% on OSWorld-Verified; today the best scores 85%, above the ~72% humans manage on the same tasks.

• Running an agent costs roughly $6-8 per hour of inference, though $3 to $15 in practice, depending on how the harness is built.

Why it's important: Enterprises are benefiting from computer-using agents in narrow, high-volume, repetitive workflows, with stable business rules and legacy interfaces or missing APIs, where success is machine-observable, failure is tolerable, and escalation is clear.

AI-NATIVE PROFESSIONAL

How to set up ChatGPT Work as a Chief of Staff

Brief: In this guide, you'll learn how to set up ChatGPT Work as a Chief of Staff that checks your connected messages, email, calendar, documents, and project trackers hourly, tracks changes, and drafts concise updates and replies.

Step-by-step:

1. Start a conversation using the starter prompt. Connect the messaging, email, calendar, and document sources relevant to your work.

2. Review the first update and flag what's important, resolved, or missing. Use the same conversation for follow-ups, meeting prep, and draft replies.

3. When a project needs its own focused working space, ask your Chief of Staff to set up a project teammate with the relevant context.

4. Leverage other follow-up prompts to adjust the hourly heartbeat, catch up quickly after time away, and prepare for your next steps.

Best practice: Connected sources and scheduled tasks depend on your plan and workspace settings. Schedule tasks in chat, not by voice.

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

BEST PRACTICE INSIGHT & CASE STUDY

Governing agentic systems in real time

Brief: Cognizant shared a guide to governing AI agents and multi-agent systems as they spread across disconnected systems and applications enterprise-wide, setting out practices for keeping them governable in real time.

Breakdown:

• Effective governance checks each interaction in real time: prompts, responses, tool calls, agent messages, before actions take effect.

• Make policy programmable, not paper-based, so risk teams can update rules centrally and every agent inherits the change at once. 

• Watch the system of agents, not the single step, and route ambiguous or high-stakes calls to human reviewers with context attached.

• As AI systems make more consequential choices, enterprises must prove who decided and under what policy, key to EU AI Act readiness.

Why it’s important: As agents become more autonomous and interconnected, trust has to be enforced at runtime, monitored across workflows, and proven after the fact. Increasingly, it will also need to be anticipated, with the control layer learning to predict incidents.

Deloitte detailed an agent action enforcement layer and an observability framework for keeping multi-agent systems under control.

Google Cloud detailed how WPP replaces guesswork with an AI-powered view of market dynamics, giving brands predictive certainty.

Deloitte published a 7-page paper on AI-driven outsourcing replacing the old model, and argued engineering is the enterprise core.

Microsoft shared how it is expanding Zero Trust for AI to assess exposure, prioritize remediation and secure AI-enabled development.

IBM argued that AI productivity leaks before it becomes enterprise value, and set out radical application development as the fix.

MIT detailed the hidden cost of AI agents in lost expertise, and why a semantic layer is now pivotal to an enterprise AI strategy.

Anthropic published a support page on watermarking Claude's text, code, and file outputs, applying the EU AI Act's rules globally.

Nvidia released Nemotron 3.5 Lightning, a small, fast model, and aims for Nemotron 4 to rival the world's best open-source models.

Meta released Muse Glimmer, a small open model that runs agents on-device, and published Zuckerberg's essay on superintelligence.

OpenAI launched GPT-5.6-Cyber, a hacking-tuned variant answering 95% of attack requests its standard model refuses, for defenders.

Anthropic introduced a Claude Code feature that allows the platform to message across sessions, so one can pick up where another left off.

Bain published a post on its global AI hackathon, where 200+ submissions from 50+ offices were narrowed to nine finalist teams.

CAREER OPPORTUNITIES

AstraZeneca - Head of Artificial Intelligence

Ode with Anthropic - AI Transformation Principal

Databricks - AI Transformation Leader

EVENTS

KPMG - CFO AI Mandate - August 18, 2026

Everest Group - Scaling AI in GBS - September 22, 2026

Microsoft Ignite - November 17-20, 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.