Infosys, the Indian IT giant, has struck a significant partnership with OpenAI to weave the American firm’s cutting-edge artificial intelligence tools, including the Codex coding assistant, into its Topaz AI platform.
The collaboration promises to revolutionise how businesses modernise software, automate routine processes, and roll out AI at scale, starting with priorities like software engineering, legacy overhauls, and DevOps.
This comes as India’s vast IT services sector navigates turbulent waters. Client spending has slowed, while generative AI surges ahead, threatening to reshape outsourcing norms. Infosys shares have tumbled more than 22 per cent this year, swept up in a wider market downturn triggered by disappointing forecasts, fears of AI eating into traditional jobs, and global jitters from the U.S.-Iran war.
As a February 2026 Gartner report highlights, 65 per cent of IT leaders view AI as both a boon and a risk to their operations.
Such alliances are becoming commonplace, as AI innovators team up with IT powerhouses to embed their tech in multinational firms. OpenAI’s earlier tie-up with HCLTech and Infosys’s recent deal with Anthropic point to this pattern. With its footprint spanning over 60 countries and more than 1,800 clients, Infosys offers OpenAI a prime route to guide companies from tentative AI pilots to enterprise-wide adoption.
The Bengaluru-based firm is ramping up its AI push. Its December quarter saw AI services deliver ₹25 billion, about 5.5 per cent of total revenue and up 30 per cent from the year before, per official filings.
OpenAI’s enterprise ambitions got a boost this week with the launch of Codex Labs, which pairs its engineers with partners such as Accenture, Capgemini, CGI, Cognizant, Infosys, PwC, and Tata Consultancy Services.
Codex, built on billions of code lines, now boasts “more than 4 million weekly active users,” according to CEO Sam Altman. Pioneering adopters cite 40 to 50 per cent quicker development times, echoing McKinsey’s forecast that AI could automate 30 per cent of software engineering by 2028.
Neither side disclosed financial details of the agreement.