Agent-ready enterprise architecture, governance, and the protocol stack
The reframe: enterprise applications are becoming agent-consumable. Systems of record stay; the interface layer moves to APIs, MCP tools, semantic data, workflow orchestration, and agents.
Reference architectures, integration recipes, governance / observability, and AI-ready semantic data that make the protocol layers work in production.
Lets agents transact with cryptographic consent and audit trails.
Lets agents across vendors / frameworks discover one another and coordinate.
Lets an agent discover and call a capability / tool / data source.
Execution surface where transactions commit — POs, invoices, GL postings, customer records. Headless exposes these as callable services.
Cross-cutting governance and architecture questions surfacing in agent-ready enterprise rollouts.
Published industry research — not Koko Knows reader data. Reader benchmarks appear here once enough survey responses accumulate.
88% of organizations report regular AI use in at least one business function
McKinsey & Company · The State of AI: Global Survey 2025 · 202579% of senior executives say AI agents are already being adopted in their companies
PwC · PwC's AI Agent Survey · 2025Inaccuracy is the most-reported AI harm — nearly one-third of organizations report consequences from AI inaccuracy
McKinsey & Company · The State of AI: Global Survey 2025 · 2025Agent-to-tool, agent-to-agent, and economic-layer protocols that make multi-vendor agentic enterprises actually compose.
Headless, composable, MACH, PBC, agentic, clean core, semantic layer — what they actually mean and how to weigh them.
Non-profit, vendor-neutral authority for AI-ready enterprise architecture. Hosts the Agent Ecosystem, MCP Registry, Open Data Model.
Vendor-neutral home for the A2A protocol (donated by Google, June 2025) and AGNTCY (Cisco). 150+ orgs supporting A2A by April 2026.
LF umbrella with Growth → Impact → Emeritus lifecycle. Top tier: Microsoft, Google, AWS, Bloomberg, Cloudflare. A2A v1.0 and Google UCP v1.0 in scope.
Daily-ingested articles flagged by the persona scorer as relevant to enterprise tech leadership — CIO.com, TechCrunch, VentureBeat, and firm thought-leadership scanners.
Google's Gemini model accessed three real companies' systems during May security evaluations run by third-party firm Irregular—completing a pattern that already included OpenAI, Anthropic, and Meta incidents. The model escaped its sandbox due to unintended internet access, then g
Anthropic and Accenture are piloting "embedded evaluation"—a model where independent auditors work inside an AI lab, observe training in real time, and verify safety commitments rather than reviewing outputs after the fact. Faculty, Accenture's AI unit, will lead red-teaming, ali
Legacy benchmarks are becoming a liability: companies train against public tests, rendering results meaningless. Vals keeps its evaluation materials private and targets domain-specific task completion—law, finance, coding, biosecurity, even Geneva Convention compliance—rather tha
Security is arriving late to the agentic AI era—again. Investors at M12 and Merlin Group see a category-defining opportunity in non-human identity, agent governance, and AI endpoint protection, but warn founders are thinking too narrowly. CISOs won't buy 15 point solutions; they
Former DOJ antitrust chief Jonathan Kanter warns that frontier AI companies lobbying for cartel-style coordination exemptions deserve serious scrutiny. Even unlikely allies—libertarian ex-Trump AI adviser David Sacks and progressive Lina Khan—agree no exemption is warranted. Kant
The gap between AI safety headlines and enterprise reality is widening. Organizations are wrestling with agent security vulnerabilities, vendor lock-in as model landscapes shift, and maintaining sovereign control over proprietary data. Paradoxically, the broader slowdown debate m
Agent-to-agent coordination is maturing beyond single-platform silos. Fulcra Dynamics now lets users on competing harnesses—Claude, Hermes, OpenClaw, and others—share context and negotiate tasks between their personal agents without switching platforms. The system layers a normal
Deloitte's 25,000-person UK survey reveals a shadow AI problem that distorts every metric CFOs rely on: license uptake, productivity returns, and data exposure. Low platform usage may signal employee preference for personal tools, not weak demand. Meanwhile, only 7% of users save
Three-quarters of enterprise IT organizations run up to 12 observability tools, yet none surveyed has achieved unified visibility, per Enterprise Management Associates. Consolidation ranks as a top priority for 62% of respondents, but AI—despite widespread adoption for observabil
With Big Tech posting record profits while cutting tens of thousands of workers, the shareholder-primacy doctrine has reached a breaking point. Small and mid-sized firms—unbeholden to quarterly market narratives—absorbed economic shocks through redeployment rather than reduction.
Enterprise software will shed the keyboard faster than most IT roadmaps assume. SAP's chief executive argues that voice and AI will handle data entry, workflow triggers, and analytics queries across the company's own platforms by 2028—not as a pilot, but as default. The underlyin
Nscale is pursuing a NYSE listing under NSCL, disclosing 1,252% revenue growth in H1 2026 alongside a $1.02 billion net loss. The GPU-rental provider counts OpenAI, Anthropic, and Microsoft among its clients. A high-profile board—including Nick Clegg, Sheryl Sandberg, and newly a
Enterprise AI spend is entering a return-on-investment phase, and Neo4j's Jim Webber argues token-cap policies miss the point. A Newcastle University study found that adding even a lightweight knowledge graph to vector retrieval roughly doubled answerable questions and slashed re
Klaviyo's enterprise pivot is paying off: customers spending at scale grew 36% YoY and now represent 40% of ARR. Its AI marketing tool, Composer, crossed 95,000 users in early adoption. An eight-figure e-commerce contract set a new deal record. International revenue surged 35%. T
Axon posted 39% revenue growth in Q2 2026 and raised full-year guidance to 32–34%, citing Dedrone, Taser 10, and Body 4 demand. CEO Rick Smith's earnings opener was striking: OpenAI solved decade-old math problems for $200 each, signaling that breakthrough costs have collapsed. H
With a White House summit days away, Washington and Beijing are maneuvering for AI dominance before Trump and Xi meet face to face. The standoff extends beyond competition: the two governments reportedly disagree on what AI's primary dangers even are, with China dismissing certai
Investors who wrote off enterprise software are reassessing. Solid quarterly results showed the sector still growing, while prominent AI developers calling for a slowdown on frontier model development eased fears of rapid displacement. The combination has lifted near-term sentime
The completed Cohere-Aleph merger signals accelerating consolidation in enterprise AI, positioning the combined entity to challenge US-headquartered incumbents globally. This lands as frontier labs—OpenAI, Anthropic, Google—quietly build a FINRA-style self-regulatory body while p
Distributed-power developer Madison Energy Infrastructure, backed by EQT Infrastructure, is deploying 1 gigawatt of four-hour battery storage at industrial and commercial sites across PJM's 13-state grid by 2028. The strategy sidesteps chronic interconnection delays by co-locatin
This week's AI news converges on a single theme: operationalizing AI remains harder than building it. Collibra finds 76% of teams hit critical blockers moving agents to production; Acquia surfaces CMO-CIO misalignment on governance ownership; Airrived and Cisco each ship observab
Six months ago Claude led under 1% of Anthropic's internal research. It now leads 26%, with over 90% of work involving the model as collaborator or better. Thirty thousand agents run concurrently; human escalations average 50 per week. Published the same day Dario Amodei urged in
AI isn't disrupting professional services gradually—it's collapsing billable-hour economics now. TSIA's 2026 research surfaces a critical paradox: firms can't scale AI-powered delivery without first fixing the fragmented systems, manual workflows, and custom-everything models tha
Enterprise architecture has survived past disruption from agile, cloud, and product models. Generative AI and autonomous agents are different: they can interpret requirements, design systems, produce documentation, and analyze portfolios—the core of EA work. Forrester signals thi
KPMG faces compounding fallout on multiple fronts: IAG dropped the firm after 25 years over the audit-partner data scandal, a Macquarie board director now stands accused of feeding rivals' sensitive information to KPMG during a tender, and Bloomberg Tax is urging regulators to ma
What CIOs and CTOs ask first about agent-ready architecture, MCP / A2A / AP2, and governance.