Walmart is introducing dedicated Agent Profiles, an Agent Gateway and operating terms for AI agents that shop on behalf of customers. The profiles, due to launch within a month, will separate agent purchases from the rest of a customer's account activity. Customers will be able to require approval for every purchase or set a spending limit, then review or delete each agent profile.

The gateway is intended to replace brittle browser-based shopping with a machine-facing connection. Walmart says it will help agents find a customer's usual products without exposing the broader purchase history to the agent. Its terms state that customers approve purchases by default, agents cannot add or change payment methods, and agents receive only information needed for the permitted task. Non-compliant agents may be blocked. The announcement matters beyond retail because it translates abstract agent governance into recognisable controls: a distinct identity, scoped permissions, transaction limits, revocation and a visible audit trail. Banks and payment providers will still need to resolve liability and authorisation standards, but Walmart's design gives the market a useful reference for how delegated digital activity can preserve customer control.


Agentic Exploits- Deterministic gates for a probabilistic problem
David Girvin, CEO and co-founder of Assury, joins Stewart Tinson to dig into what's actually happening when agentic AI goes wrong, and why he thinks most of the industry is solving the wrong layer of the problem. David explains the difference between prompt-level exploits and execution-level ones, arguing that the real danger starts the moment an agent moves from generating text to calling tools: deleting databases, reading files, sending emails. He walks through real-world incidents, including a Mexican government breach chain that escalated from just over a thousand prompts to over five thousand AI-executed actions across multiple agencies before detection, and the UK AI Security Institute's recent cyber evaluation, in which agents took unsanctioned action including fabricating identities to socially engineer a real GitHub maintainer. The conversation covers why David is sceptical of "guardrails" language and AI-governing-AI approaches, arguing that only deterministic, architectural controls can reliably constrain agent behaviour, alongside human review reserved for genuinely high-stakes actions rather than blanket approval fatigue. He breaks down credential starvation, session risk accumulation, and why classifier-based tools keep failing inconsistently on identical actions, pointing to a named frontier lab's own zero trust paper as an example of the industry misjudging what actually works. Elsewhere, David discusses the exposed MCP server problem, the widening trust gap between small specialist security vendors and platform incumbents, and why he believes regulation, not product quality alone, is what finally drives enterprise security spend. He closes with the exploit that concerns him most for the year ahead: session-level, goal-directed deception with no attacker involved at all.
Share this post
The link has been copied!