Welcome executives and professionals. The enterprise AI leaderboard has sparked significant debate recently, but who is the true market leader? It's complicated.

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.

Note: Starting next week, Enterprise AI Executive will be published twice weekly, on Wednesdays and Sundays at the usual time.

In today’s briefing:

• a16z’s state of enterprise AI.

• 2026 Process Optimization Report.

• Introducing OpenAI Frontier.

• Build-buy-borrow strategy.

• Transformation and technology in the news.

• Insights for Executive+ members.

• Career opportunities & events.

Read time: 4 minutes.

MARKET INSIGHT

a16z’s state of enterprise AI

Brief: VC firm Andreessen Horowitz (a16z) surveyed 100 VPs and C-level executives across the Global 2000. The sample included 88% of firms with $1B+ in revenue, and 30% exceed $10B annually.

Breakdown:

• OpenAI remains the enterprise LLM leader, with 78% using its models in production. Anthropic reached 44%, overtaking Google Gemini.

• OpenAI still controls the largest LLM wallet share at roughly 56%, though Anthropic (17%) and Gemini (16%) continue to gain steadily.

• Data shows a shift toward 3rd-party apps over LLMs. Even in areas like knowledge management, where in-house builds historically dominated.

• Microsoft dominates enterprise AI with M365 Copilot, alongside incumbents GitHub Copilot, Agentforce, and ServiceNow.

• Enterprise LLM spend rose from roughly $4.5M to $7M over two years and is projected to grow another 65% this year to $11.6M.

Why it’s important: Predictions that slower model progress and competition would flatten the field have proven wrong. A small group of providers continues to consolidate share and accelerate innovation. And in a enterprise AI where current perception can drive future reality, the stakes couldn’t be higher.

PROCESS INTELLIGENCE

2026 Process Optimization Report

Brief: Celonis, a global leader in Process Intelligence, published its 34-page Process Optimization Report based on a survey of 1,649 business leaders, revealing a gap between agentic AI ambitions and operational readiness.

Breakdown:

• 85% of firms aim to become an "agentic enterprise" within three years, yet 76% admit their current processes are holding them back.

• AI agents require optimized, AI-ready processes and the process data and operational context that comes from process intelligence.

• Without both, AI agents can’t understand how a business actually runs. 82% of leaders believe AI will fail to deliver ROI without this foundation.

• The top barriers to agentic AI adoption are limited internal expertise (47%) and difficulty getting AI to understand business context (45%).

• 58% report their departments still do not operate seamlessly together, preventing the end-to-end visibility required for effective Enterprise AI.

Why it’s important: Process Intelligence provides the "common language" that allows AI agents to understand how work flows across departments and systems, identify friction points, and navigate the complex reality of business to execute actions that drive business outcomes.

ENTERPRISE ACCELERATOR

Introducing OpenAI Frontier

Brief: OpenAI introduced Frontier, a new enterprise platform for building, deploying, and managing AI agents. It onboards agents and equips them with shared context, learning, and permissions to operate safely at scale.

Breakdown:

• Business Context connects enterprise systems (e.g. data warehouses) so AI agents operate on the same information as teams.

• Agent Execution lets AI agents work in parallel to complete complex tasks reliably across production environments.

• Built-in evaluation and optimization loops show what works and what fails, allowing agents to improve over time.

• Enterprise security and governance are built in with permissions, and auditable actions to ensure agents act safely.

• HP, Oracle, and State Farm are among the first adopters, with OAI embedding engineers on-site to help get agents into production.

Why it’s important: Enterprises are under pressure to scale AI results, but progress is constrained by how agents are designed, governed, and deployed. Frontier shifts the focus from model capability to operational execution, helping organizations turn AI agents into reliable systems of work.

BEST PRACTICE INSIGHT

Build-buy-borrow strategy

Brief: KPMG published a 17-page report to help leaders make the right agentic AI investments based on value, risk, and readiness. The real risk isn’t moving too fast, it’s choosing the wrong model.

Breakdown:

• Agentic AI is hitting operational reality, yet most firms remain stuck at a key decision point: whether to build, buy, or borrow.

• Build works best when differentiation, control, and sovereignty matter, and firms have mature engineering teams and long-term intent.

• Buy is ideal for speed and proven capabilities, especially when vendor solutions meet needs and internal AI maturity is limited.

• Borrow suits firms prioritizing flexibility and rapid scale, enabling co-creation with acceleration partners while sharing risk.

• The report details considerations for each model across strategic fit, customization, cost, risk, and more (image above).

Why it’s important: Agentic AI can redefine what’s possible, but value depends on choosing the right operating model. Applying a structured build-buy-borrow framework helps leaders align investments with strategy and risk tolerance, and ensure initiatives drive enterprise-wide impact.

Goldman Sachs outlined its AI outlook, detailing near-term expectations and how adoption is shaping priorities and decisions in boardrooms.

EY published guidance for boards leading in an AI-shaped world, emphasizing non-automatable accountability, and key questions.

Google Cloud COO Francis deSouza shared lessons from scaling enterprise AI, including how to define and launch a first high-impact pilot.

Deloitte examined how intelligent agents are reshaping product innovation, shifting organizations from incremental improvement to disruption.

Microsoft surveyed 500 enterprise leaders across $1B–$50B firms, identifying five practices that enable successful AI agent deployments.

The International Safety Report 2026 delivers a 221-page assessment of AI capabilities, risks, and governance, informed by over 100 experts.

OpenAI released GPT-5.3-Codex, a flagship coding model combining advanced programming and reasoning, while powering its own training process.

Anthropic launched Claude Opus 4.6, its most capable model yet, adding multi-agent Claude Code, a large context window, and Office integrations.

Perplexity introduced Model Council, running prompts across multiple AI models simultaneously and synthesizing responses into one unified answer.

OpenAI signed a $200M agreement with Snowflake, enabling enterprise customers to access GPT-5.2 for agent development on proprietary data.

SpaceX acquired xAI, merging into one integrated engine valued near $1.25T. The core logic of the merger focuses on Orbital Data Centers.

ElevenLabs raised a $500M funding round, valuing the AI voice company at $11B and tripling its valuation within just twelve months.

CAREER OPPORTUNITIES

OpenAI - Global Partnerships Executive

JPMorgan - AI Transformation Executive Director

Nasdaq - Enterprise AI Senior Director

EVENTS

ElevenLabs Summit - 11th February, 2026

Chief AI Officer Summit - 25th February, 2026

CDAO 2026 - February 25-26, 2026

66% of Enterprise AI Executive subscribers are C-level or VP/Director, from firms such as Microsoft, Deloitte, Goldman Sachs, and much of the Fortune 500.

Multi-edition 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.