THEBUSINESSBYTES BUREAU
MUMBAI, NOVEMBER 20, 2026
Four out of five Indian businesses (80 per cent) experienced at least one
artificial intelligence (AI) failure in the past year, with 44 per cent
reporting financial losses as a result, highlighting growing risks amid the
rapid adoption of AI technologies, according to a new study by BSI.
Despite the increasing exposure to AI-related risks, less than half (46
per cent) of Indian organisations have established processes to manage AI
incidents and ensure timely responses, pointing to significant gaps in
corporate preparedness and risk management.
The findings are part of BSI's global study, The Cost of AI Failure,
which surveyed 1,287 business leaders across nine countries, including India.
The research examined instances where AI systems deviated from their intended
objectives or produced harmful outcomes.
Besides financial losses, 43 per cent of Indian respondents reported
safety implications arising from AI failures, while 38 per cent cited loss of
productive time and 33 per cent flagged regulatory or compliance-related
challenges.
The potential financial exposure could increase significantly with the
growing deployment of agentic AI systems. Separate economic modelling
commissioned by BSI and conducted by Alma Economics estimates that failures
involving agentic AI could put up to 3.6 per cent of annual market revenue at
risk in certain sectors.
The estimate reflects potential market-level exposure under modelled
scenarios and does not represent actual losses suffered by Indian businesses.
The findings assume significance as Indian companies increasingly
integrate agentic AI into their operations, with 24 per cent of respondents
already deploying the technology at scale and another 26 per cent using it for
specific applications.
However, concerns over emerging risks remain pronounced, with 58 per cent
of businesses identifying security as a major concern and 44 per cent pointing
to inadequate AI governance.
The study also highlights a growing gap between organisations' confidence
in their AI safeguards and their actual implementation. While 81 per cent of
respondents expressed confidence in their organisations' safe AI policies, only
63 per cent said such policies were in place.
Similarly, although 86 per cent believed their organisations were
transparent about AI data and algorithms, only 39 per cent understood the data
being used to train or deploy these systems.
On the governance front, however, Indian businesses have made progress.
The proportion of organisations with established AI governance programmes rose
to 42 per cent from 24 per cent a year ago, while those having risk assessment
approaches increased to 57 per cent from 30 per cent during the same period.
Commenting on the findings, Emmanuel Herve, Managing Director for South
Asia and Southeast Asia (SSEA) at BSI, said as AI reshapes businesses,
economies and society, Indian leaders must pay greater attention to the
consequences of AI failures.
AI failures are becoming increasingly common and costly, making it
essential for organisations to prevent such incidents wherever possible and
openly share lessons when they occur, he said.
Realising the benefits of AI requires responsibility, robust governance and effective safeguards, while early action can help reduce financial and legal risks and strengthen trust, Herve added.
Despite the growing concerns, business confidence in AI remains strong, with 89 per cent of Indian organisations expecting their AI investments to increase.