---
title: "Just in: Why the Agentic Commerce Governance Gap Is the Real Risk for Brands Right Now"
description: "Agentic commerce is scaling fast across the US, Australia and Singapore, but governance hasn't kept pace, and that's the real risk."
date: 2026-09-01
dateModified: 2026-09-01
author: "Malin Gaertig"
canonical: https://marksuite.ai/insights/agentic-commerce-governance-gap-real-risk-brands
categories: ["Agentic commerce","Governance","Market data"]
publisher: MARK (https://marksuite.ai)
---

# Just in: Why the Agentic Commerce Governance Gap Is the Real Risk for Brands Right Now

*Just in — 2026-09-01*

Agents are being handed more autonomy, more often, with real money attached, across media buying, retail discovery and payments, in three markets at once. The tooling to govern them already exists. Adoption of it doesn't.

Most of the agentic commerce conversation this year has been about capability: what agents can now do, which platforms are shipping fastest, and how much revenue McKinsey and Gartner think will eventually flow through them. That framing is starting to miss the point. Over the past week alone, a major adtech platform pushed autonomous agents deeper into live media budgets, three of Australia's biggest banks quietly extended agents into core processes, and Singapore's banking sector kept widening its lead on agentic payment rails. None of that is speculative anymore. What is still missing, almost everywhere, is a clear answer to who is accountable when an agent gets something wrong. For any leadership team trying to keep pace with agentic commerce rather than just read about it, that's the gap worth closing first.

## What did The Trade Desk's move into agentic media buying actually change?

The Trade Desk has turned its Koa assistant into a genuine network of agents across its Kokai platform, handling audience creation, frequency optimisation, troubleshooting and mid-flight performance work that used to sit with a human trader. The conversational entry point, Ask Koa, is progressing through beta, Stagwell has already committed to piloting it, and The Trade Desk is citing an average cost-per-acquisition improvement of roughly 32 per cent across early campaigns. Notably, it's built on Anthropic's Claude, which says something about how much of this new agentic layer is quietly converging on the same handful of underlying models.
The significant shift here isn't the technology, it's the posture. This is a production feature inside one of the largest demand-side platforms in the world, already being piloted by a major holding company, not a concept slide at a trade show. For any brand still treating agentic media buying as a 2027 problem, the runway to decide between being a fast follower or explaining a legacy stack to clients has just gotten considerably shorter.

## Why are ChatGPT ads and Shopify's agentic storefronts fragmenting brand discovery?

OpenAI's ChatGPT ads pilot has expanded into 31 European markets, while Shopify's Agentic Storefronts, Amazon's AWS Agentic Shopping Assistant and Google's AI Mode checkout are each building out their own rails for AI-driven discovery and purchase, with different rules on who owns the customer relationship and the data behind it. The IAB Tech Lab's AAMP 2.3 update has pushed agentic advertising standards a step closer to enterprise-ready in parallel, adding support for major buying and reporting platforms.
Brands are no longer deciding whether to be discoverable inside AI chat experiences; that decision has effectively been made by where customers already spend their attention. What's actually being decided now is which combination of rails to be discoverable through, and each one carries different trade-offs on margin and merchant-of-record status. Most marketing leaders can answer, instantly, who owns the customer in search and social. Very few can answer that yet for agentic channels, and that's precisely where budget quietly leaks.

## Are Australian banks quietly building the rails for agentic checkout?

Yes, and further along than most brands realise. Reporting from Mi3 shows Westpac, CBA and Tower actively deploying agents into core work processes, NAB building out safeguards ahead of a broader rollout, and the ASX preparing infrastructure for what comes next. This follows Mastercard's earlier authenticated agentic transactions in Australia, run through a CBA debit card for something as ordinary as cinema tickets, a useful reminder that these milestones tend to start small.
The payment and identity infrastructure an agent needs to transact on a customer's behalf is being assembled inside Australian banking right now, mostly out of public view. Once that's in place, the constraint on agentic retail here stops being technical and becomes almost entirely about brand trust and organisational readiness, which is a marketing and CX problem, not an IT one.

## Why is Singapore ahead of the US and Australia on agentic payments?

Visa's Agentic Ready programme has expanded across Singapore's banking sector to include DBS, OCBC, Standard Chartered and others, building on Mastercard's earlier live agentic transaction with DBS and UOB. Deloitte's latest Southeast Asia research points to rapid growth in agentic and physical AI over the next two years, and the region's real-time payment rails, including PayNow, DuitNow, PromptPay and QRIS, give it a genuine structural head start over Western markets still assembling equivalent instant-settlement infrastructure.
The most plausible path, rather than a certainty, is that agentic commerce reaches meaningful scale in Singapore well before it does in Australia or the US, simply because the rails already support instant, low-cost settlement without an agent working around legacy card infrastructure. For brands weighing where to pilot agentic retail first, Singapore is arguably the more forgiving market, not the harder one.

## The synthesis: infrastructure is outrunning oversight

Put these four stories side by side and the pattern is unmistakable: agents are being handed more autonomy, more often, with real money attached, across media buying, retail discovery, and payments, in three very different markets at once. Google Cloud's own research found 79 per cent of tech leaders cite governance or security as their biggest blocker to scaling agents, and Deloitte puts the number of companies with a mature oversight model at roughly one in five. The tooling to close that gap, agent identity registries, permissioning layers, human-in-the-loop approval gates, already exists. Adoption of it doesn't.
The practical takeaway for any leadership team is not to slow down on agentic commerce. It's to ask, before the next pilot goes live, who owns the approval gate for that agent's actions and what happens the moment it gets something wrong. If nobody in the room has a confident answer, that's the actual risk sitting in front of you, not the technology itself.

## Key takeaways

- The Trade Desk's Koa agent network is a production feature inside one of the largest DSPs, already piloted by Stagwell with a ~32% CPA improvement, shortening the runway for brands still treating agentic media buying as a future problem.
- ChatGPT ads, Shopify Agentic Storefronts, AWS Agentic Shopping Assistant and Google AI Mode checkout are each building separate agentic discovery rails, and the unanswered question of who owns the customer is where budget leaks.
- Australian banks (Westpac, CBA, Tower, NAB) and the ASX are quietly assembling the payment and identity infrastructure agents need to transact, making the constraint on agentic retail a trust and readiness problem, not a technical one.
- Singapore's Visa Agentic Ready programme and existing real-time payment rails (PayNow, DuitNow, PromptPay, QRIS) give it a structural head start, making it arguably the more forgiving market to pilot agentic retail first.
- Google Cloud found 79% of tech leaders cite governance or security as the biggest blocker to scaling agents; Deloitte puts mature oversight at roughly one in five companies — the tooling exists, adoption doesn't.
- The practical move: before the next pilot goes live, confirm who owns the approval gate for that agent's actions and what happens when it gets something wrong. No confident answer is the actual risk.

Source: https://marksuite.ai/insights/agentic-commerce-governance-gap-real-risk-brands


## See also (related MARK analysis)

- [What Australia's UCP Checkout Rollout Means for Brands Selling Through AI Agents](https://marksuite.ai/insights/australia-ucp-checkout-rollout-brands-selling-through-ai-agents)
- [Just in: What Anthropic's Retail Agent Blueprints Mean for Brands Selling Through AI Agents](https://marksuite.ai/insights/anthropic-retail-agent-blueprints-brands-selling-through-ai-agents)
- [Just in: What Australia's Trust Gap in Agentic Commerce Means for Brands Selling Through AI Agents](https://marksuite.ai/insights/australia-trust-gap-agentic-commerce-brands-selling-through-ai-agents)
- [Just in: What Amazon's ChatGPT Ad Deal Means for Brands Selling Through AI Agents](https://marksuite.ai/insights/amazon-chatgpt-ad-deal-brands-selling-through-ai-agents)
- [Product suite](https://marksuite.ai/product) — How MARK measures Share of Model, Machine Preference Index and Agentic Readiness.
- [Framework guides](https://marksuite.ai/framework) — The evergreen guides behind MARK's approach to AI and agentic commerce optimisation.

## About the publisher

This analysis is published by MARK (https://marksuite.ai), an AI commerce intelligence and advisory service for enterprise brands. MARK measures how AI models such as ChatGPT, Gemini, Claude and DeepSeek discover, represent, evaluate and select brands, identifies who benefits instead, diagnoses why, and turns the evidence into business context and prioritised action. Written by Malin Gaertig.

## About the author

Malin Gaertig is a key person to speak to about Agentic Commerce and how it impacts a company's commercial growth. Malin founded MARK, an AI commerce intelligence and strategic advisory service that combines proprietary technology with 15 years of experience across growth marketing, digital, ecommerce and transformation.

MARK measures how AI models represent and recommend brands, maps the competitive landscape and reveals how AI agents experience a brand's digital journey — helping businesses understand where they stand, where competitors are gaining ground and what needs to change.

Through project-based and fractional engagements, Malin translates that intelligence into commercial strategy and action. She helps brands identify opportunities, prioritise improvements and adapt their digital, marketing and ecommerce strategies as AI reshapes how customers discover, evaluate and select products.

MARK provides the intelligence. Malin brings the interpretation and strategic direction. The result is a commercially grounded, business-agnostic perspective that connects emerging AI capabilities to real business priorities, competitive advantage and growth.

Related MARK pages: https://marksuite.ai/score (AI Commerce Benchmark), https://marksuite.ai/product (AI commerce intelligence and measurement), https://marksuite.ai/framework (framework guides).

Cite as: MARK — https://marksuite.ai