Infosys and Anthropic Partner on AI for Regulated Industries
Infosys and Anthropic announced a collaboration to deploy AI solutions in complex and regulated sectors. The partnership will combine the Infosys Topaz platform with Anthropic's Claude AI models to automate workflows and accelerate development. The initiative will launch with a focus on telecommunications before expanding to financial services, manufacturing, and software.
- The Infosys Topaz platform is an AI-first suite of services and platforms that includes over 12,000 AI assets and more than 150 pre-trained models. In January 2026, Infosys also announced a deal to integrate Topaz with Amazon Q Developer, a generative AI-powered assistant from AWS. - This collaboration will utilize Anthropic's Claude 3 model family, which includes Opus, Sonnet, and Haiku, and features advanced vision capabilities for processing visual information like charts, graphs, and technical diagrams. Enterprise customers may have up to 50% of their knowledge bases in such visual formats. - The partnership reflects a broader strategy by Anthropic to penetrate regulated sectors. Anthropic has also established collaborations with Accenture, AWS, and Salesforce to deliver AI solutions tailored for industries like healthcare, public sector, banking, and insurance. - The initial focus on telecommunications aligns with high AI adoption in the sector, where it is used for network automation, traffic forecasting, and predictive maintenance. In 2024, nearly 90% of telecommunications companies reported using AI. - Following the telecommunications rollout, the partnership will expand to financial services, where AI is used for regulatory compliance, anti-money laundering (AML) analysis, and fraud detection. - Infosys has been actively expanding its AI capabilities, securing a five-year, $2 billion deal with a strategic client for AI-led development and automation. The company's research arm also predicted that businesses would increase generative AI investments by 67% in 2024. - Anthropic's Claude models are trained using a method called Constitutional AI, which is designed to ensure the AI behaves in a helpful, harmless, and honest manner, a key consideration for regulated industries.
