Your ERP just hired its first AI employee
Something shifted in Dynamics 365 this month, and it’s worth paying attention to even if you’re not the type to get excited about software updates.
Microsoft’s 2026 Release Wave 1 is live. The headline change: autonomous AI agents are now embedded across the Dynamics 365 stack. Not the chatbot kind. Not the “here’s a suggestion” kind. These are software agents that do actual work — research leads, process purchase orders, schedule field technicians, triage support tickets — without someone sitting in front of a screen telling them what to do.
Microsoft’s own framing: “AI is no longer assistive. It is operational.” That’s a big claim. Let’s look at what’s actually shipping.
What the agents do
In Business Central, AI agents now handle sales and purchase scenarios end to end. Think: a purchase order that gets created, reviewed against inventory levels, matched to a supplier, and routed for approval — with a human only stepping in for exceptions. For companies running lean finance and ops teams, this is where the time savings start to compound.
In Sales, there’s a qualification agent that researches inbound leads before your sales reps do. It pulls data from your CRM, enriches it from external sources, and scores the lead. Your reps get a brief instead of a blank record. For teams managing high lead volumes, that’s the difference between calling twenty cold leads and calling five warm ones.
In the Contact Centre, containment rates go up because the AI handles more first-touch interactions across chat, email, and voice. Supervisors get real-time insights across every channel without toggling between dashboards.
In Field Service, a Scheduling Operations Agent allocates technicians based on skills, location, urgency, and availability. Less dispatching overhead, more time on the tools.
In Commerce, the updates are about B2B complexity — multi-outlet ordering, attribute-based pricing, cross-legal-entity inventory lookup. These are the features that matter when you’re running five warehouses across three legal entities and your customers want to see stock availability before they order.
What the agents don’t do
They don’t replace people. This is important, and not just as a talking point. Every agent operates within defined parameters. They execute workflows, but a human sets the guardrails. They make decisions, but those decisions are logged, auditable, and reversible. If an agent approves a purchase order above a threshold you didn’t intend, you’ll see it in the audit trail.
They also don’t fix bad data. This is the part that doesn’t make it into Microsoft’s press releases. An AI agent that processes purchase orders is only as good as the product catalogue it reads from. A lead qualification agent is only as useful as the CRM data it enriches against. If your master data is inconsistent, incomplete, or stale, the agents will automate your problems faster than your team currently creates them.
Why data quality is the real prerequisite
Every Dynamics 365 implementation Braintree has done over the past twenty years comes back to the same truth: the technology is never the bottleneck. The data is.
Before you turn on autonomous agents, you need to know your data is clean. Product records are current. Customer records aren’t duplicated. Pricing rules are accurate. Inventory counts reflect reality. If you can’t trust your data with a human reading it, you definitely can’t trust an AI agent acting on it.
This isn’t a reason to wait. It’s a reason to prepare. Run a data quality audit. Fix the gaps. Then deploy the agents on a foundation that actually supports them.
What this means for South African businesses on Dynamics 365
If you’re already on Business Central or any Dynamics 365 application, these agents are coming to your environment. They’ll roll out over the next few months as part of the standard release cycle. You don’t need a new licence for them — they’re included in your existing D365 subscription.
The question isn’t whether the agents will be available. It’s whether your environment is ready for them. The organisations that get the most out of this wave will be the ones that did the groundwork: clean data, defined workflows, clear approval hierarchies.
Braintree helps companies across South Africa get their Dynamics 365 environments ready for what’s next. Whether that’s a data quality review, a Wave 1 readiness assessment, or a conversation about what agentic AI means for your specific workflows — we’re here.
If you’re running Business Central and want to understand what these agents mean for your operations, get in touch.