Edition in partnership with
Welcome executives and professionals. As rivals access similar AI capabilities, how can enterprises turn agentic investment into unique, defensible advantage?
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:
Winning on the agentic frontier.
Reshaping AI partner channels for vendors.
Anthropic’s Fable 5.2 takes the lead.
Redesigning management for agentic AI.
Transformation and technology.
Insights for Executive+ members.
Career opportunities & events.
Read time: 4 minutes.
Note: Enterprise AI Executive is published on Wednesdays only. We no longer publish on Sundays.
MARKET & BEST PRACTICE INSIGHT
Image source: IBM
Brief: IBM’s 31-page survey of 1,000 C-suite executives, with Microsoft, finds every respondent is increasing agentic AI budgets, but those strong across six capabilities report 6x the productivity gains and twice the AI revenue.
Breakdown:
Its Agentic AI Value Model spans ecosystem strategy, governance, controls, work design and execution, connected by a learning loop.
Firms with a high deployment rate and strong foundations (above) report 13x the ROI of those with a low rate and weak foundations.
Only 22% report meaningful governance design and operational controls, and none deploy them across the entire agentic lifecycle.
63% say scaling needs an ecosystem strategy but only 22% have one, and just 11% report transformational redesign of workflows.
Why it's important: Lasting agentic advantage comes from capabilities rivals can’t buy, not from the agents themselves. Enterprises should measure outcomes, build learning loops, keep a governed agent registry, and scale workflows once the capabilities to support them are ready.
IN PARTNERSHIP WITH TELEPORT
Brief: Running agents in production takes more than a sandbox. Now generally available, Teleport Beams is a trusted runtime that gives agents delegated identity, controls infrastructure access, and makes their actions visible and auditable.
With Teleport Beams, you can:
Connect to infrastructure without secrets
Approve plans before execution
Flag behavior that drifts from intent
Stop agents and their connections
BEST PRACTICE INSIGHT
Image source: McKinsey & Company
Brief: McKinsey detailed how AI is reshaping not just what technology enterprises buy but how vendors sell it, as partners drive about 60% of vendor revenue on average, yet their channels still run on a traditional SaaS playbook.
Breakdown:
Over half of large AI projects now rely on partner-led POCs to prove technical credibility and win executive buy-in to scale them.
Illustrative data (above) shows vendors widening partnerships across four archetypes since 2020–22, led by cloud and AI platforms.
Vendors should make five channel moves, first mapping partners to the four archetypes and engaging them from codesign through adoption.
The other three are a dedicated AI partner program, AI-driven channel operations, and partners supplying forward deployed engineers.
Why it’s important: Many AI proofs of concept never reach production, and partners are often key to whether they do. Vendors should reward partners for adoption, not sales, while enterprises should involve partners early and expect them to supply forward deployed engineers.
INNOVATION INSIGHT
Image source: Artificial Analysis
Brief: Anthropic released Claude Fable 5.2 and Mythos 5.2, faster upgrades to its most powerful AI models with per-token prices cut by 20%, as Fable 5.2 takes the No. 1 spot on Artificial Analysis’ Intelligence Index leaderboard.
Breakdown:
Fable 5.2 overtook Google’s limited-release Gemini 4 Carbon, though Anthropic still advises starting most work on its cheaper Opus 5.5.
Mythos is limited to vetted security and life sciences groups, while Fable is the same model with stronger cyber and bio guardrails.
1,000+ Fable agents built an operating system in 8 days and formalized Fermat’s Last Theorem in 30 hours, as Opus 5.5 completed 26%.
Fable 5.2 costs twice Opus 5.5 per token and 20% less than 5.1, but uses more tokens, so max-effort tasks cost about double 5.1’s.
Why it’s important: The frontier keeps moving fast, but cost per task matters more than price per token. Enterprises should match models to the work, using Opus for most tasks and Fable for the hardest ones, and track total spend as agents take on longer, more complex jobs.
AI-NATIVE PROFESSIONAL
Brief: In this guide, you’ll learn how to use ChatGPT to explain KPI movements, turning dashboards, metric definitions, exports, and campaign notes into a source-backed brief that separates drivers from hypotheses.
Step-by-step:
State the KPI, comparison period, expected direction, and the decision the analysis supports, then attach the dashboard and exports.
Add the metric definition, segment cuts, and campaign context, then ask ChatGPT to check data quality and suggest useful breakdowns.
Run the starter prompt and review confirmed drivers separately from hypotheses, with charts, caveats, open questions, and sources.
Use the follow-up prompt to test each driver, then validate actions with the metric owner before changing a dashboard or process.
Best practice: Start with a stable metric definition and trusted source data, and never treat segment correlation as proof of cause.
For the full guide, including prompts, upgrade to Executive+ or The Boardroom.
BEST PRACTICE INSIGHT
Image source: Deloitte
Brief: Deloitte argues management is one of the most important roles to redesign for AI, because while AI makes information more accessible, it raises the value of the judgment, accountability, and orchestration managers provide.
Breakdown:
As AI makes information accessible across all layers of the organization, the human judgment of managers becomes more important.
Deloitte highlights six shifts in managers’ roles spanning information, expertise, decisions, learning, orchestration and change.
AI changes the work, but managers still connect strategy to execution, resolve trade-offs, build capability and own the outcomes.
Firms can redesign management (see image above) without choosing between keeping the status quo and cutting layers indiscriminately.
Why it’s important: High-performing organizations will redesign management around outcomes, decisions, and orchestration, rather than traditional supervisory roles. Organizations that simply remove layers risk creating a gap between strategy and execution.
McKinsey detailed how CFOs can make AI pay off by setting measurable targets, redesigning processes end to end and leading change.
Strategy& argued that IT defines financial firms’ business models, with a five-stage framework to test systems against strategy.
TCS published a 31-page study of 1,333 leaders, finding most see productivity as AI’s main value, but just 9% run agentic workflows.
Forrester found no imminent AI bubble burst, though firms funding the build-out face strain, and explored Tencent’s agent scaling.
Accenture’s CTO discussed where firms should start with AI, as 70% of 300 CIOs surveyed say AI mainly aids individual productivity.
HCLTech outlined an AI-native software delivery model with human checkpoints, and launched a 48-question sovereign cloud assessment.
a16z argued patching can’t keep up with machine-speed attacks, so defenses must shield unpatched systems as exploits get cheaper.
IMD asked how much judgment boards have ceded to AI, encouraging them to check how far it shapes their views and how they’d spot errors.
Anthropic rolled out Claude Sheets and Whiteboards in beta, made Claude Data generally available, and announced Claude Meetings.
OpenAI introduced Businesses in ChatGPT, a beta hub where brands claim verified profiles, track mentions and set up shopping agents.
The Hackett Group released XT Version 2, an AI platform pairing its benchmarks with client data to speed up transformation roadmaps..
Nvidia’s board ordered a review of reports CEO Jensen Huang was targeted by Chinese intelligence, as he admitted an affair to staff.
Anthropic opened IPO investor pitches, and said in a new filing it expects coming models to cross its AI research safety threshold.
Palantir reported Q3 revenue rose 96% to $2.31B, as its US commercial sales of $984M topped US government sales for the first time.
Google announced TPU 8tx and 8i Max, specialized AI chips due next year, and cut its Gemini Omni 1.2 Flash video model’s price 30%.
Chinese models now reportedly take 89% of US enterprise spend on open-source AI models, led by Moonshot AI’s Kimi with a 61% share.
Access Executive AI Index: The top 391 AI playbooks, by industry and function, with direct links to each. Updated weekly.
Get the extended version of Enterprise AI Executive.
Unlock the full AI-native professional guides.
CAREER OPPORTUNITIES
OpenAI - Executive in Residence, Cyber
Mars - Global Head of AI & Data
Anthropic - Head of Technology Partnerships
EVENTS
Microsoft Ignite - November 17-20, 2026
HFS - The AI Change Crisis - November 26, 2026
Gartner - Agent Orchestration - December 16, 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.
Lewis, Ashley, Mark
