AT&T is sharpening its artificial intelligence strategy with the launch of a network-based digital receptionist designed to protect customers from robocalls and fraud while enabling more seamless communication.
The US-based telecom giant, which already blocks or labels over 2 billion robocalls each month, is now testing agentic AI technology that screens incoming calls in real time — without requiring app downloads or draining phone batteries, Andy MarkusChief Data Officer, AT&T, said in a blog post.
U.S. consumers received just over 4.1 billion robocalls in August 2025, a 4 percent drop from July and 6.1 percent lower than August 2024, according to YouMail’s Robocall Index. This marks the fourth straight monthly decline since volumes peaked in April, leaving robocall activity about 16 percent below that high.
Despite the recent slowdown, the first eight months of 2025 recorded 36.7 billion robocalls, up roughly 7.1 percent compared with the same period in 2024. August’s activity averaged 133.9 million calls per day, or about 1,550 per second, down from July’s 144 million daily average and 1,666 calls per second.
Fighting Spam with Agentic AI
The AI receptionist acts like a human assistant, answering calls, asking questions such as “Who’s calling?” and “What is this regarding?” and then deciding whether to connect, block, or take a message. Using multiple large language models (LLMs) and advanced speech-to-speech processing, it can detect suspicious patterns, disconnect fraudulent calls, and even provide live transcripts for users to monitor.
Unlike traditional spam filters, the AT&T tool is integrated directly into the network and linked to a customer’s phone number, ensuring it works even when the device is out of service range. Trusted contacts can be placed on a “Do Not Screen” list, ensuring important calls always go through.
Future AI Capabilities
AT&T envisions a broader AI roadmap beyond spam defense. The same agentic AI framework could eventually make restaurant reservations, handle scheduling, or manage calls customers can’t take in the moment, all guided by real-time text prompts and customer preferences.
Privacy and Security Built In
The company emphasizes strict security policies, stating that data is only used to tailor experiences and is kept private and protected. By blending decades of fraud-prevention expertise with cutting-edge AI, AT&T aims to deliver faster, more reliable call screening while maintaining customer trust.
With this AI-powered digital receptionist, AT&T is positioning itself at the forefront of telecom innovation—leveraging AI not just to fight robocalls but to redefine how users interact with their phones, setting a new benchmark for customer-centric AI services.
AT&T is also expanding its AI strategy beyond call screening by exploring multiple agentic AI use cases designed to enhance customer convenience and network efficiency. The company is testing tools that can autonomously manage tasks such as making reservations, handling scheduling, and managing calls when users are unavailable.
AT&T is also leveraging decades of expertise in fraud detection to develop advanced algorithms that integrate with large language models for faster, more accurate decision-making. These initiatives are built directly into the network to ensure seamless operation without extra apps or battery drain, reflecting AT&T’s broader goal of delivering secure, privacy-focused AI experiences that simplify daily communications and strengthen customer trust.
Baburajan Kizhakedath
