IT stocks: Anthropic AI sparks ‘DeepSeek 2.0’ moment; selloff an opportunity for investors?

Nirmal Bang Institutional Equities, in a fresh note, described Wednesday’s selloff in Infosys, Tata Consultancy Services Ltd (TCS) and other IT stocks as a “DeepSeek 2.0 moment”. The brokerage said the selloff was an overreaction but also a wake-up call for Indian IT firms. It said Indian IT stocks were hit by global contagion risk rather than any domestic deterioration, adding that the move reflected a spillover from the sharp derating in US enterprise software and consulting stocks, not a reaction to Indian company-specific fundamentals.

“For Indian IT, this is not an earnings cliff but a strategy wake-up moment. Nifty IT fell 5.9 per cent despite stable near-term revenue visibility. Total contract value has remained almost flat since the first quarter of FY25, indicating that AI rhetoric has not yet translated into larger or longer-duration deals. We believe the conclusion is clear. AI built as a wrapper will not defend growth or margins. Indian IT firms must either build AI from the ground up or acquire core model capabilities,” it said.

Nirmal Bang said Palantir and Anthropic had accelerated the re-pricing by attacking time and workflow economics. It said Palantir’s claim that SAP ECC to S 4 migrations can be completed in weeks rather than years directly compresses duration, labour intensity and client lock-in.

“Anthropic’s Claude co-worker plugins showed how quickly foundation model providers can ship end-to-end workflows across legal, sales and analytics. We note that when execution moves to the model layer, wrapper software and services lose insulation almost instantly,” it said.

The brokerage said Wall Street had been uneasy about software stocks for months, but sentiment shifted sharply over the past forty-eight hours. The fear, it said, is that AI could collapse entire layers of the software value chain, which led to indiscriminate selling across SaaS, consulting and data analytics stocks.

Nirmal Bang believes the selloff creates opportunity rather than terminal risk.

It said what makes the current debate more uncomfortable for Indian IT services companies is that the data has not yet validated the optimism. Total contract value disclosures for Tier 1 Indian IT companies have remained flat, or close to flat, since the first quarter of FY25, with no sequential step-up that would normally signal the start of a new investment cycle. This matters, it said, because TCV is a forward indicator.

“Management commentary has leaned heavily on AI-led deal momentum, yet the aggregate numbers tell a more restrained story. We believe this reflects two forces working simultaneously. The first is macro caution. US enterprises remain conservative on discretionary spending amid economic uncertainty, political noise and an unclear rate trajectory. Decision-making cycles are longer. Scope sizes are smaller. Clients are choosing containment over transformation. AI has not yet been enough to override that caution,” it said.

The second force, according to the brokerage, is structural. It said AI remains powerful but unreliable in complex enterprise settings, with error rates still high, hallucinations unresolved and governance frameworks immature.

“Indian IT companies, lacking proprietary large language models, are largely operating as integrators of third-party intelligence. What is being built, in most cases, is a wrapper around APIs from OpenAI or Anthropic. This limits both differentiation and client confidence. Events around Anthropic’s recent releases underline this risk. When a foundation model provider ships a native tool that replicates workflow-level functionality, entire products built as wrappers can be rendered redundant overnight,” it said.

Nirmal Bang said this dynamic is now being internalised by CIOs, which is slowing commitment, compressing deal sizes and explaining why TCV has not inflected despite rising AI rhetoric.