From Generic Assistants to Embedded AI Workflow Agents
AI security workflows and AI advertising tools are shifting from broad, one-size-fits-all chatbots to specialised agents wired into day-to-day operations. Two recent moves underline this transition. ServiceNow is expanding its platform with the acquisition of Armis and the launch of a Dispute Management AI Agent with Xactly, signalling a deeper push into security and revenue workflows. In parallel, StackAdapt has released a Model Context Protocol (MCP) Server that exposes its campaign intelligence AI directly inside tools like Claude and other AI environments. Instead of requiring users to live inside a single platform, these approaches embed AI agents where security analysts and marketers actually work. The result is a quieter but more profound shift: AI becomes part of the workflow fabric, not a separate destination, promising faster incident handling, more responsive campaign optimisation and fewer hours lost to manual reporting or spreadsheet gymnastics.
ServiceNow, Armis and AI Security Workflows
ServiceNow’s acquisition of Armis extends its reach from classic IT service management into broader AI security workflows that span physical, operational and cyber assets. Armis is positioned around cyber exposure management across operational technology, Internet of Things and other physical assets, an area where security buyers increasingly want unified visibility. By pulling this real-time asset intelligence into ServiceNow, security teams can track connected devices, OT systems and cloud resources from a single workflow platform. In practice, this promises faster incident triage, streamlined risk assessments and fewer disconnected tools for security operations centres. Instead of hopping between consoles to understand exposure, analysts can lean on AI agents embedded in ServiceNow to prioritise alerts, correlate issues across domains and trigger automated responses. The strategic bet is that integrated AI agents will turn ServiceNow into a central nervous system for enterprise security, not just a ticketing front-end.
AI Agents in Revenue and Dispute Workflows
Beyond security, ServiceNow is also targeting revenue operations with its Dispute Management AI Agent developed in partnership with Xactly. This agent plugs AI directly into cross-platform revenue workflows, particularly around compensation and commercial disputes between sales and finance. The goal is to move dispute handling from email chains and manual review into structured, AI-assisted processes. ServiceNow has already signalled a focus on agent-based workflows for areas such as manufacturing orders, warranty claims and configuration; the Xactly collaboration extends that logic into compensation and sales disputes. For revenue leaders, the promise is faster resolution cycles, clearer audit trails and closer alignment between sales, finance and operations. As enterprises pilot these domain-specific AI agents, adoption patterns will reveal which use cases deliver real productivity gains and which remain experimental, shaping how platforms like ServiceNow compete with rivals that also blend workflow tools and AI agents.
StackAdapt’s MCP Server and Campaign Intelligence AI
On the advertising side, StackAdapt’s MCP Server reshapes how campaign intelligence AI shows up in marketers’ daily work. Instead of forcing teams to log into a single platform, the StackAdapt MCP Server connects its Ivy AI marketing assistant to AI tools they already use, including large language models, agents and workflow automation systems. Campaign performance, configuration details and creative assets can be queried conversationally in tools like Claude, covering channels such as CTV, display, native, audio, DOOH and programmatic linear TV. Marketers can monitor pacing, audit creative and analyse audience-level results in real time without manual reporting or spreadsheet wrangling. The MCP Server also lays groundwork for agent-assisted optimisation, where AI systems continuously watch campaign metrics, surface insights and trigger actions within user-defined guardrails. This approach turns campaign data into a live, embedded resource, rather than a static dashboard that teams check only after problems appear.
Productivity Gains and the Challenges of Embedded AI
Taken together, ServiceNow AI agents and the StackAdapt MCP Server illustrate a broader shift toward embedded, domain-specific AI. Security teams gain unified views across physical, operational and cyber assets, enabling quicker incident resolution and more coherent risk management. Revenue operations benefit from automated dispute handling that shortens resolution cycles and clarifies ownership. Marketers get campaign intelligence AI woven directly into their preferred tools, allowing always-on optimisation instead of periodic, manual reviews. Yet integration complexity and oversight remain critical challenges. ServiceNow must successfully blend Armis’ operational technology and IoT capabilities and Xactly’s data into its platform without creating implementation headaches for large customers. StackAdapt must ensure its open-web intelligence fits smoothly into diverse AI ecosystems. Organisations adopting these tools will also face a learning curve as teams learn to trust, supervise and fine-tune AI decisions, balancing automation with human judgment.