AT&T is broadening its agentic AI capabilities with a new internal platform that lets employees design their own autonomous assistants. According to a blog by Andy Markus, AT&T’s Chief Data Officer, the company has added a drag-and-drop agent builder—Ask AT&T Workflows—that automates multi-step processes across customer operations, network engineering, and internal support functions. The tool extends AT&T’s secured GenAI environment, which includes Ask Docs and Ask Data, so employees can build task-oriented agents using authenticated enterprise information.
The first production deployment, developed jointly by AT&T Business and the technology development teams, now processes customer service update requests by synchronizing data and installing information across systems in real time. Every workflow includes mandatory human checkpoints, full action logging, and strict role-based access controls. Markus notes that AT&T’s recent performance on AI industry benchmarks—Spider 2.0 text-to-SQL accuracy and the GSMA Open-Telco LLM leaderboard—signals that the company is not only consuming AI but increasingly building its own agentic tools.
AT&T is also extending these capabilities to network operations. Engineers can assemble agents that correlate telemetry, check recent change logs, identify known issues, open tickets, propose fixes, and compile incident summaries. These workflows aim to reduce outage resolution time while keeping human oversight embedded throughout the process. AT&T frames this shift as moving AI from “information” to “action,” enabling employees to offload repetitive tasks and focus on higher-value problem solving.
• Ask AT&T Workflows introduces a graphical tool for building multi-step AI agents
• First in-production workflow automates customer service update requests
• Agents integrate securely with Ask Docs and Ask Data for enterprise-specific insights
• Network engineering agents can perform telemetry correlation and propose resolutions
• All workflows enforce human approval points, data isolation, and retention policies
“We’re centered on one word: agentic… AI agents are fulfilling the promise of what AI can do for companies,” said Andy Markus, Chief Data Officer, AT&T.
🌐 Analysis
AT&T’s investment in agentic automation aligns with broader telco efforts to embed AI assistants directly into OSS, BSS, and network-operations cycles. The approach is conceptually similar to initiatives at Mplify, where new frameworks aim to define cross-organizational agent-to-agent communication for service orchestration. Other carriers—including Verizon, Deutsche Telekom, and NTT—are piloting AI-augmented workflows, but AT&T’s internal builder gives it a unified method for scaling agents across both business and network domains.
