WebMCP could turn the web into an agent-native interface
WebMCP gives websites a structured way to expose actions directly to AI agents, creating the potential for a machine-readable interaction layer alongside the human web experience.
Latest thoughts
MCP could become the connective tissue of the AI-native GTM stack
Model Context Protocol could shift AI from another application in the GTM stack to a governed interface capable of discovering, coordinating and acting across business systems.
Team Brain suggests a different business software model
Team Brain's public proposition suggests an AI-native workspace could reduce software fragmentation by assembling data, agents, integrations and compute around business outcomes rather than individual applications.
SpaceX is turning the AI stack into an industrial stack
SpaceX is assembling control across energy, chips, compute, models and applications, creating an AI strategy in which infrastructure may matter as much as the models it produces.
When agents can pay, autonomy becomes economic
Amazon Bedrock AgentCore payments moves payment execution inside the agent workflow, turning spending into a governed machine action alongside reasoning, tool use and execution.
AI governance is becoming an identity problem
As AI agents begin acting across enterprise systems, identity becomes the mechanism that connects delegated authority, permission controls, auditability and accountable autonomy at runtime for organisations.
Stripe is building the economic control plane for AI
Stripe's reported OpenRouter acquisition suggests a broader ambition to connect AI revenue, usage metering, model routing and settlement into a programmable economic infrastructure layer.
Everything is a plugin
DeepSeek Harness points to a wider shift in AI architecture, where the model becomes one replaceable component inside a composable runtime of tools, memory, execution and control.
Software is becoming agent-addressable
Cloudflare Radar Researcher offers an early view of software designed for AI agents, where machine-readable tools, APIs and semantics sit alongside conventional interfaces.
Human-in-the-Loop Is Not a Safety Strategy
Human approval can become slow, superficial and unreliable as autonomous systems scale, making bounded autonomy, consequence thresholds and targeted intervention a stronger enterprise control model.
The rise of programmable trust
Trust is becoming an operational capability as enterprises connect identity, authority, runtime controls, evidence and human accountability across increasingly automated digital systems.
Policies are not governance
AI policies define intent, but effective governance requires technical controls, workflow enforcement and evidence that can constrain system behaviour, expose exceptions and enable accountable intervention.
AgentOps: The enterprise discipline that does not exist yet
As AI agents gain authority across enterprise systems, organisations need AgentOps to connect ownership, identity, permissions, observability and intervention without allowing operational control to slow useful innovation.
The next GTM stack will be built for agents
As AI agents take on more revenue work, organisations will need a governed control environment connecting customer data, organisational knowledge, execution tools, permissions and measurable feedback.
Microsoft's Harvey deployment changes the legal AI question
Microsoft's adoption of Harvey signals that legal AI is moving into enterprise workflows, changing how legal work, professional judgement, governance and accountability must be organised.
The next customer journey is assembled, not browsed
AI agents are compressing discovery, comparison and purchasing, requiring brands to combine human appeal with structured product truth, operational access and governed customer relationships at scale.
TikTok Shop or Shopify? Decide the role before you choose the platform
TikTok Shop and Shopify solve different ecommerce problems. This article explains how UK businesses should compare discovery, control, margin, data, operations and platform risk before deciding which role each should play.
The Next Enterprise Operating System Will Not Look Like Software
Enterprise software is shifting from visible applications to orchestration layers that coordinate workflows, data, people and AI agents, moving operational control and strategic advantage behind the interface.
AI transformation looks more like cloud transformation than software deployment
AI may arrive through familiar applications, but enterprise value depends on redesigning workflows, governance, skills and management disciplines around the technology.
AI strategy is becoming an operational discipline
AI strategy is shifting from identifying promising use cases to managing the workflows, ownership, controls and performance disciplines required to turn experimentation into repeatable enterprise value.
Enterprise AI has a visibility problem
As AI moves into enterprise workflows, organisations need visibility across identity, information retrieval, tool use, policy decisions and outcomes to preserve trust, control and accountability.
Why governance is becoming runtime infrastructure
As AI systems move from generating outputs to taking actions, governance must become an executable operational layer that can enforce policy, manage authority and generate evidence while work is happening.
The Real Enterprise AI Race Is Not About Models
As leading AI models become easier to access, lasting enterprise advantage will depend on workflow control, proprietary context, orchestration, governance and the discipline to turn intelligence into measurable operating performance.
AI pilots are easy. Operational trust is the hard part
AI pilots can prove technical capability quickly, but enterprise scale depends on operational trust: clear boundaries, observable behaviour, accountable ownership and evidence that systems remain controlled in production.
Why most AI strategies fail before the technology does
Enterprise AI strategies often break down because unclear workflows, fragmented information and weak governance prevent capable technology from producing reliable, controlled and measurable business value.
The CIO is becoming the governor of autonomous systems
As AI systems gain authority to make decisions and initiate work, CIOs must govern the identities, permissions, orchestration and evidence that keep enterprise autonomy accountable.
The enterprise AI moat is moving below the model
As AI moves from generating answers to executing work, durable enterprise advantage is shifting towards the infrastructure that governs context, identity, workflows, risk and performance.
Beyond the Job Description: Designing Work for the Agentic Organisation
AI is dissolving the boundaries of traditional roles. In the agentic organisation, work flows dynamically between humans and machines — replacing static job descriptions with fluid responsibility frameworks built around outcomes.
When AI Becomes the First Responder: The New Front Door to Every Brand
As AI systems increasingly become the first point of contact between customers and brands, conversational design, trust, and escalation strategy are becoming core business capabilities rather than technical features.
Why AI Won't Replace Workers — But It Will Replace Jobs
AI rarely replaces entire occupations. Instead, it absorbs tasks until traditional roles hollow out from within — leaving workers employed but fundamentally redefining what their jobs mean.
From Roles to Responsibilities: Why AI Is Forcing Organisations to Redesign Work
AI is dissolving traditional job structures by fragmenting tasks and redistributing work between humans and machines. The organisations that succeed will redesign roles around responsibility, system oversight, and outcomes rather than static job descriptions.
From Acceleration to Accountability: The Metrics That Will Define AI Success in 2026
As AI shifts from experimentation to operational infrastructure, organisations must replace vanity metrics with outcome-driven measurement frameworks linking AI to financial performance, operational efficiency, workforce productivity, and risk governance.
When the AI Stack Becomes the Problem: Why Tool Consolidation Is Now Strategy
Enterprise AI programmes are increasingly collapsing under fragmented tool stacks. The organisations that succeed will move from collecting tools to designing unified AI platforms that enable scale, governance and velocity.
The Quiet AI Divide: Why Some Organisations Will Pull Away in 2026
AI is not creating a level playing field. Organisations with the right data, governance, and learning structures are quietly building compounding advantage, while others remain trapped in endless experimentation.
When AI Becomes Invisible: Planning for the Era Where Intelligence Is Just Infrastructure
AI's ultimate success is not dramatic disruption but quiet ubiquity. As artificial intelligence fades into the background of enterprise systems, the organisations that thrive will be those that treat it as infrastructure rather than innovation.
The Second-Order Effects of AI: Why Automation Is the Least Interesting Part of the Story
AI's biggest impact is not automation. It is the delayed shift in power, trust, careers and culture that unfolds 12-36 months later. Leaders who measure only efficiency will miss the structural changes that determine long-term success.
The End of AI Optimism: Realism Becomes the Competitive Advantage in 2026
Enterprise AI has entered its realism phase. With 95% of pilots failing to deliver ROI and only 14% of CFOs seeing clear impact, disciplined cost control, governance, and measurable outcomes now separate durable advantage from expensive experimentation.
The Talent Illusion: Why Hiring Won't Fix Your AI Capability Gap
Hiring elite AI talent won't fix stalled transformation. The real constraint is organisational design, decision rights, data access, and incentives determine whether intelligence becomes capability or frustration.
Why 2026 Strategy Is Built Around Systems, Not Projects
By 2026, AI exposes the limits of project-based strategy. Organisations that shift to owning and governing living systems will compound value, while those still “delivering” AI will watch it decay.
Agents Without Context: Why Fragmented Data Is the Real Threat to Autonomous Systems
Autonomous agents fail not because they lack intelligence, but because they operate on fragmented enterprise truth. Unified, real-time, policy-governed context is now the prerequisite for safe and scalable autonomy.
From Pipelines to Platforms: Why AI Demands a New Data Operating Model
AI exposes the limits of batch-era data pipelines. Sustained decision quality in volatile environments requires closed-loop, platform-based data architectures, not greener dashboards.
AI Risk Is Not IT Risk: Why Boards Need a New Oversight Model
AI systems now shape enterprise decisions, not just infrastructure. Boards that fail to treat AI risk as enterprise risk, with structured, visible oversight, will face regulatory, reputational and strategic consequences they cannot delegate away.
The Auditability Gap: Why AI Decisions Can't Be Explained After the Fact
AI systems are moving into regulated, high-impact roles faster than most organisations can explain or defend them. The next competitive advantage belongs to those who design auditability into AI from the start, because confidence is no longer enough.
The New Moat: Why Speed Beats Accuracy in AI-Driven Markets
In AI-driven markets, competitive advantage belongs to organisations that iterate fastest, not those that launch the most accurate models. By 2026, learning speed becomes the new moat.
From Guardrails to Governance: Why Prompt Rules Aren't Enough
Prompt rules shape outputs. Governance defines responsibility. As AI systems become autonomous, enterprises must move from configuration-based guardrails to architectural accountability, or risk scaling liability instead of value.
From Pilots to Permanence: Why 2026 Ends the Era of AI Experiments
By 2026, pilot culture is no longer sufficient. Organisations must industrialise AI with governance, ownership and workflow redesign, or risk stagnation as experimentation turns into avoidance.
The Great AI Split: Why 2026 Is the Point of No Return
By 2026, AI advantage shifts from tool access to institutional memory. Early adopters who embedded AI in operations are compounding learning and cost advantages that late movers cannot easily replicate, creating a structural split in enterprise performance.
Who Owns the Agent? Designing Accountability for Autonomous Systems
Agentic AI is accelerating faster than governance structures can adapt. Enterprises must move beyond shared oversight and define clear, lifecycle ownership for autonomous systems, or risk accountability diffusion at scale.
Who Owns the Agent? The Accountability Gap in Autonomous Systems
As AI systems evolve into autonomous agents, responsibility fragments across teams while accountability remains unclear. The organisations that win in 2026 will be those that treat agents as governed actors, with named owners, clear oversight, and structured accountability.
The Committee Trap: When AI Governance Becomes the Bottleneck
AI governance is expanding rapidly, but committee-heavy oversight often slows transformation and increases shadow risk. The future belongs to organisations that replace gates with guardrails and embed governance directly into their AI operating systems.
The Leadership Lag: When AI Moves Weekly and Boards Move Quarterly
AI is moving faster than traditional oversight. Discover why continuous governance, real-time visibility and adaptive leadership now define enterprise advantage.