Anthropic and Infosys Partner on AI Agent Development
Anthropic has entered into a partnership with global consulting firm Infosys to develop AI agents, according to a recent industry roundup. The collaboration signals a push to apply Anthropic's models to enterprise-level agentic systems. This partnership could create new demand for data tailored to specific business and industry use cases for AI agents.
- The partnership will integrate Anthropic's Claude models with Infosys Topaz, an AI-first set of services and platforms, to develop agentic AI for industries like telecommunications, financial services, and manufacturing. A primary focus is on modernizing legacy systems by using AI agents to accelerate migration and reduce costs. - Anthropic's "Constitutional AI" is a key technique for training models like Claude to be helpful and harmless, using a set of principles to self-correct responses, which can reduce the need for extensive human labeling. This method, also known as Reinforcement Learning from AI Feedback (RLAIF), has been shown to produce models that are more robust against jailbreaking attempts compared to Reinforcement Learning from Human Feedback (RLHF) alone. - Evaluating agentic systems requires a shift from traditional AI benchmarks to assessing multi-step reasoning, tool use, and error recovery. Key metrics include task success rates, cost per task, latency, and the accuracy of tool selection, moving beyond simple output correctness to measure behavioral reliability. - While synthetic data offers scalability and speed, human labeling remains crucial for tasks requiring nuance, contextual understanding, and bias mitigation. Hybrid approaches are often most effective; models trained primarily on synthetic data see significant performance improvements when fine-tuned with even small amounts of human-labeled data. - Go-to-market strategies for AI infrastructure startups are shifting from a "growth at all costs" mindset to a focus on capital efficiency and verifiable customer outcomes. Key performance indicators now include Return on AI Investment (ROAI), which measures revenue generated from automated workflows against the cost of model inference and computes. - The fundraising climate for AI companies is robust, with AI capturing nearly 50% of all global venture funding in 2025, a significant increase from 34% in 2024. Foundation model developers like Anthropic and OpenAI raised a substantial portion of this capital, with investors showing massive interest in AI-linked opportunities.
