Amdocs outlined an artificial-intelligence-led growth plan at the Goldman Sachs Communacopia + Technology Conference on Tuesday, September 8, 2026, as it seeks to extend its agentic operating system, aOS, across its telecom customer base and into additional regulated industries.
CEO Shimie Hortig said Amdocs serves more than 400 service providers in more than 90 countries, with systems that affect about 3 billion customers daily. The company said it currently captures roughly one-third of customer IT spend in core business support systems and operations support systems domains. Amdocs reported trailing-twelve-month revenue of $4.65 billion, a dividend yield of 3.8%, and has raised its dividend for 14 consecutive years, while its shares are down 24% year to date. InvestingPro rates its financial health as GOOD.
Hortig described a four-pillar strategy: expand aOS adoption among existing telecom customers; enter one additional regulated vertical beyond telecom over the midterm to long term, potentially with acquisitions; develop two to three new agentic-era products for global commercialization; and continue internal transformation so Amdocs can act as customer zero. The company said 10 customers are already using aOS.
The Liberty Latin America deal, announced about a month earlier in August, was cited as a marquee example. Liberty Latin America handed over its entire IT organization to Amdocs, which will manage and transform the operation using aOS to reduce costs and accelerate product launches. Hortig said the company's challenge is that AI technology is improving quickly, but customer adoption remains limited, and Amdocs needs more visible wins to prove the model.
Amdocs said its business model has remained outcome-based for the company's 40-year history, with compensation tied to meeting key performance indicators and service-level agreements rather than time and materials. Internally, the company has spent more than a year implementing an agentic software development life cycle, shifting team structures toward more builders and fewer enablers, and flattening management layers. Hortig said context, knowledge and know-how are essential for AI models to support mission-critical systems, and Amdocs plans to demonstrate results through broader aOS adoption and additional customer wins.
