Infosys Founder N.R. Narayana Murthy Critiques Misleading AI Practices

Infosys founder N.R. Narayana Murthy did not hold back as he directed his criticism towards the technology sector in India, targeting the “trend” of categorizing regular software as artificial intelligence (AI).

Speaking at TiE Con Mumbai 2025, Murthy cautioned that many offerings, which are showcased as AI, are simply “outdated programs” adorned with exaggeration.

Addressing a crowded gathering at TiE Con Mumbai 2025, Infosys founder N.R. Narayana Murthy exhibited considerable doubt about how AI is being portrayed in India. He remarked, “It seems to have become a trend in India to brand everything as AI. Many standard programs are being promoted as AI.”

Murthy proceeded to elucidate the two fundamental concepts of AI, cutting through the misconceptions. “Firstly, machine learning involves large-scale correlations that assist in making predictions based on extensive data,” he stated.

“This is simple machine learning. The second concept is referred to as deep learning. Deep learning mimics the functioning of the human brain,” he added, clarifying the essential elements that set authentic AI apart from basic programming.

Expanding on deep learning, Murthy indicated that it can handle “unsupervised algorithms,” in contrast to machine learning, which is predominantly data-supervised. “Whereas machine learning mainly deals with supervised algorithms that require substantial data, deep learning uses data to form new branches of programs or create new conditions. It can then make decisions,” he elaborated.

“Thus, unsupervised data, which employs deep learning and neural networks, possesses significantly more potential to emulate human behaviours effectively. However, what is being labeled as AI today consists of trivial, outdated programs,” Murthy asserted, shattering the misconceptions surrounding AI in India’s technology landscape.

Meanwhile, Infosys is developing its own AI models, concentrating on Small Language Models (SLM) that utilise both open-source elements and proprietary datasets. The organisation is actively promoting Generative AI (Gen AI) for targeted industry applications and use cases, striving to transcend mere superficial AI labels to create impactful tools.