Edition in partnership with

Welcome executives and professionals. Leaders, boards and investors repeatedly ask similar questions: Is software becoming easier to replicate? Will AI-native challengers disrupt incumbents?

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:

• How AI is reshaping software moats.

• McKinsey on scaling AI inference.

• Mistral challenges China's open models.

• Accenture sees broadening AI demand.

• Transformation and technology.

• Insights for Executive+ members.

• Career opportunities & events.

Read time: 4 minutes.

Note: To take Enterprise AI Executive to the next level, starting October 14, we’ll move to publishing once a week, every Wednesday. The number of Executive+ insights in each edition will double.

MARKET & BEST PRACTICE INSIGHT

How AI is reshaping software moats

Brief: EY, drawing on insights from 400 CTOs, revealed how AI is reshaping software moats, engineering talent and economics, and why the advantage lies in the data, workflows and trust that surround the code itself.

Breakdown:

• More than four in ten tech leaders say customers routinely ask at renewal whether parts of their product could be rebuilt or replaced with AI.

• CRM apps can hit 80% feature parity in weeks, but without the underlying data, regulatory logic, customizations they are empty shells.

• One in three CTOs surveyed expect over half of new production code to be AI-authored or agent-generated within three to five years.

• AI already drives software revenue, with over a third of respondents linking over 25% of their software revenue to AI capabilities.

Why it’s important: Software's moat has moved from code to the context around it: data, workflows, integrations and trust. As AI removes friction from coding, validation and governance become the real constraints, and talent scarcity shifts from coding capacity to judgment.

IN PARTNERSHIP WITH RESOLVE

How to make every IT practitioner an expert

Brief: Your best IT experts shouldn't be the only ones who can solve critical issues. Learn how to capture tribal knowledge, turn proven procedures into governed workflows, and help every practitioner resolve issues like an SME.

What you’ll learn:

• Prioritize frequent, repeatable procedures that deliver the fastest returns

• Turn expert-led incident procedures into tested, governed workflows

• Measure impact through MTTR, ticket volume, onboarding hours saved

• Spread critical knowledge from a few specialists across the whole team

Join the live webinar on October 20.

MARKET & BEST PRACTICE INSIGHT

McKinsey on scaling AI inference

Brief: McKinsey shared how the shift from AI training to inference is reshaping AI economics, infrastructure, and investment priorities across the semiconductor ecosystem, as enterprises strive to scale AI.

Breakdown:

• For years, most AI compute went to training, driven by GPU scarcity, frontier model races, and ever larger clusters to train those models.

• By 2030, inference is expected to account for about 60% of AI demand, with training falling to about 40% of the total workload mix.

• The industry is moving from building AI intelligence to serving it at scale, a different engineering challenge and cost structure.

• Training runs on homogeneous clusters, while inference is heterogeneous in hardware and constraints: latency, throughput, and cost.

Why it's important: C-suite leaders now ask whether AI creates value at an acceptable cost to serve. Costs arrive immediately, measurable productivity gains take longer, and costs grow as agents do more work. How fast those costs fall will shape where and how quickly enterprises scale AI.

INNOVATION INSIGHT

Mistral challenges China's open models

Brief: France's Mistral AI introduced Mistral Large 4, a 1-trillion-parameter open-weight model nicknamed "Le Chonk," which outperforms every non-Chinese open system available today by a “substantial margin.”

Breakdown:

• China's Kimi K3 still leads on coding at 68% (image above), but Le Chonk's 62% beats the 44% of Beam, Reflection AI's US open model.

• Mistral claims a top-5 global ranking in cybersecurity and 82% on a bug test that Claude Opus 5.5 and GPT-6 Astra mostly refused.

• On legal work, Mistral says tester Vals scored it nearly 3x GPT-6 Astra on Harvey's benchmark, ahead of every Chinese open model.

• Cyber experts and government agencies get a less-filtered version in a three-week preview, with the open weights due on October 27.

Why it’s important: Open-weight models help enterprises and governments run AI on their own infrastructure, keeping sensitive data in-house and reducing token costs. Until now, the strongest options have been Chinese, but Beam and Large 4 offer credible Western alternatives.

AI-NATIVE PROFESSIONAL

How to diagnose a stalled deal with ChatGPT

Brief: In this guide, you'll learn how to give ChatGPT deal stage history, closed activities, call transcripts, emails, deal threads, security or procurement notes, and account context, then ask it to explain the blocker and next action.

Step-by-step:

1. Define the deal, time window, and decision to support, then attach stage history, activity, transcripts, emails, and account context.

2. Ask ChatGPT to build a clear timeline of the deal before it names a root cause, so the diagnosis stays grounded in what happened.

3. Run the starter prompt and review the evidence, prior attempts, dependencies, escalation path, and recommended next move.

4. Once the blocker is agreed, use the follow-up prompt to create a focused recovery plan with owners, dependencies, and proof points.

Best practice: Keep the diagnosis factual and time-bound. If the blocker is unclear, list the smallest check that distinguishes between the competing explanations.

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

MARKET INSIGHT

Accenture sees broadening AI demand

Brief: Accenture closed fiscal 2026 with fourth-quarter revenue of $18.7B, up 7% in local currency and above guidance, as AI-led reinventions, record managed services bookings and its partner ecosystem drove broad-based growth.

Breakdown:

• New bookings reached $22.2B with a 1.2 book-to-bill, led by a record $12.8B in managed services and 37 clients booking over $100M.

• Nearly 100 clients began their first generative, agentic or physical AI work with Accenture in Q4, taking the FY26 total past 400.

• Bookings with eight emerging AI partners like Anthropic, OpenAI and Palantir more than tripled in FY26, and revenue more than doubled.

• FY27 guidance of 3-6% growth leans on 2-2.5% from acquisitions, with lower Q4 pricing and intense competition flagged as headwinds.

Why it’s important: Large-scale reinventions, many driven by AI, are driving strong demand. These span transforming functions and building out digital cores. Clients increasingly believe AI will help them achieve more across the enterprise, but they remain at very different stages of readiness.

Deloitte explored how frontier AI incidents show agents can bypass intended boundaries, with four questions to govern agent access.

McKinsey argued that capturing the larger value of agentic AI will require leaders to rethink the shape of the organization itself.

TIAA released its Retirement in the Age of AI and GLP-1s Survey: 40% see workplace AI as a threat to their retirement plans.

AWS explained why executive sponsorship drives transformation success, and launched two Forward Deployed Engineer partner pathways.

Microsoft shared an 8-page paper on sovereign AI with NVIDIA for regulated sectors, and showed how its EY alliance scaled AI value.

Anthropic argued GLM-5.3 was released without meaningful safeguards to limit misuse, and asked people what they really want from AI.

Anthropic introduced Claude Code mods, add-ons that customise how it looks or behaves, such as a check that blocks risky deletions.

OpenAI is testing a ChatGPT desktop Meetings plugin that turns meetings into tailored notes and actions, with Enterprise due soon.

Anthropic launched a beta Claude sidebar in Google Docs, Sheets and Slides for paid plans that can edit files or ask approval first.

Cohere launched North 2, an upgraded agentic platform with tighter governance and cost controls, and released its Embed 5 models.

Reflection AI introduced Beam, a 501B-parameter open-weight model for coding and agents that it says rivals leading Chinese models.

Elon Musk hinted at renaming SpaceXAI to "SpaceXSI" after the U.S. push to rename the technology, saying "we will make that change."

CAREER OPPORTUNITIES

Citi - Head of AI Customer Experience

OpenAI - Partner Director, HCL, Wipro & Cognizant

3M - Senior Director, Enterprise AI Transformation

EVENTS

OxGen AI Summit - October 15-16, 2026

Oracle AI World - October 25-28, 2026

Snowflake Expedition - November 3-5, 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.