Amperity Highlights Trusted Customer Context as Critical to Managing Agentic AI Risk
Author: Belinda Lloyd, Customer Success Lead, Amperity
As autonomous AI agents gain traction across Asia Pacific, Amperity says trusted customer context will be critical to managing agentic AI risk and ensuring systems act with the right identity, permissions, and governance.
Agentic AI is moving quickly across Asia Pacific. BCG found that 77% of workers say their businesses are experimenting with or deploying autonomous agents. However, only 33% understand them well. Deloitte expects adoption among the region’s consumer businesses to rise from 29% to 76% within two years.
That gap matters because agentic AI risk grows as AI systems gain the ability to take action. Agents can now do more than generate content or recommend a next step. They can update records, choose audiences, trigger journeys, adjust offers, and coordinate work across multiple systems.
The moment AI starts taking action, the risk changes with it.
A person can review a wrong answer. However, a wrong action may already have reached a customer, changed a record or triggered another workflow. If an agent works from fragmented identity, stale preferences or inconsistent consent, it can make the wrong decision with remarkable confidence. Worse, it can execute that decision at machine speed.
Risk Escalates When Agentic AI Starts Taking Action
Most organisations do not have one clean, universally accepted version of a customer. Email, point of sale, loyalty, customer service and the data warehouse may each hold a different record. These systems may also apply different matching rules or update at different speeds.
People have learned to work around those gaps. Autonomous systems will simply act on the information they receive.
Consider a common retail scenario. An agent needs to decide which customers should receive an offer. One system shows a high-value customer. Another shows an unsubscribed email address. Meanwhile, a third system has not yet registered yesterday’s return.
If the business cannot resolve those signals into the right customer context, agentic AI risk increases. The agent might suppress a valuable relationship, recommend a product the customer just returned, or contact someone through a channel they have opted out of.
The problem is not necessarily that the model lacks intelligence. Instead, it lacks a complete, current, and governed picture of the person affected by its decision. Giving that system more autonomy only makes the gap more consequential.
Managing Agentic AI Risk Starts With Trusted Customer Context
Trusted context does not mean every data point must be perfect. Instead, the organisation needs to know who the customer is, which signals are current and which permissions apply. It should also be able to explain why the system made a decision.
In addition, the context needs to fit the decision at hand. A marketing offer, service intervention, and financial decision should not rely on identical assumptions or access rules.
An agent cannot reason well about a customer it cannot recognise. Customer intent can also change from one moment to the next. Moreover, not every useful piece of data should be available for every action.
Identity, real-time signals and governance can turn a static customer record into live operating context.
Trusted customer context is not simply preparation for agentic AI. It is what an agent uses to decide whether to act, what action is appropriate, and when it should stop or ask for help.
Three Questions Every Agentic AI Workflow Should Answer for Governance
If a team cannot clearly answer the following questions, the workflow is not ready for greater autonomy.
- What does the agent know? Teams should be able to identify the customer and business context informing the decision. They should know where that information came from, how recently it changed, and whether the agent has permission to use it. An apparently complete profile offers little value if the underlying identity is wrong or an important consent signal is missing.
- What is the agent allowed to do? Access to information should not automatically grant authority to act. Agents need explicit boundaries around the systems they can reach, the actions they can take, and the conditions that require human approval. The same least-privilege principle used for employees and applications should apply to AI.
- Can the organisation see and correct the outcome? Teams need a record of the context used, the decision made, the action taken, and the downstream result. High-impact actions should be reversible. People also need a clear way to intervene when an outcome falls outside policy or customer expectations.
These tests extend established risk and security principles. NIST places governance across the AI lifecycle, while OWASP highlights agentic risks including identity and privilege abuse, tool misuse, and cascading failures. The technology is new, but the need for controlled access and accountable decisions is not.
Reducing Agentic AI Risk Means Setting Clear Boundaries
The most capable agent is not necessarily the one that takes the most actions. Instead, it should recognise when the context is strong enough to act, when it has reached a boundary, and when a person needs to step in.
Asia Pacific has no shortage of momentum around agentic AI. The next competitive advantage will come from turning that momentum into trusted action.
Organisations that connect live customer context with clear permissions, visible decisions and measurable outcomes can move quickly without losing control.
Before businesses give AI more agency, they need confidence in the context guiding it. At machine speed, trust cannot be a manual checkpoint added after the decision. It has to be part of every decision the system makes.
About Belinda Lloyd, Customer Success Lead, Amperity
Bel Lloyd is a contributor in the Australian Martech landscape and the Customer Success Lead at Amperity, where she operates at the intersection of people, technology, and data. Recognised as the 2023 Digital Marketer of the Year (Women in Digital), Bel drives impactful growth for enterprise brands by unifying complex customer data to deliver personalised experiences and measurable business outcomes, including leading the charge in closed-loop attribution having facilitated a first-of-its-kind industry partnership with MixIn by Endeavour Group and Criteo to bridge the gap between digital advertising and verified in-store sales.
Beyond her technical leadership, Bel is a dedicated advocate for Women in Tech. She serves as a Lead Mentor in the She Codes community, empowering women to transition into technical roles, and is a frequent industry speaker on the role of diversity in fueling innovation. Through her advocacy and technical expertise, Bel continues to champion inclusion and share forward-thinking perspectives on the future of the digital economy.
About Amperity
Amperity is the AI-powered Customer Context Platform that helps brands turn fragmented customer data into trusted customer context. In an AI-first world, brands need more than data. They need a complete, real-time understanding of their customers. Amperity connects customer signals, resolves identity, and powers continuous decisioning so brands can deliver more relevant experiences in the moments that matter. More than 400 brands worldwide rely on Amperity, including Alaska Airlines, DICK’S Sporting Goods, BECU, Virgin Atlantic, and Wyndham Hotels & Resorts. Founded in 2016, the company operates globally with offices in Seattle, New York City, London, Argentina and Melbourne. Learn more at amperity.com.