New Delhi: Indian companies are spending more on artificial intelligence than ever before, but a new Adobe report released today suggests most of that money is getting stuck somewhere between the pilot and the production line. Only 6.7 per cent of the 475 Indian firms surveyed have scaled AI across multiple business workflows, while 59 per cent are struggling to move beyond pilot projects, according to the India findings of the Adobe 2026 State of AI Readiness Report.
The numbers tell an uncomfortable story for corporate boardrooms. Investment is clearly not the problem. As many as 77 per cent of companies say they are increasing their AI spending year on year, and almost two in three feel they are still underinvesting. Yet around 44 per cent of respondents admitted they are either working on their very first AI use case or have quietly shelved a project after an unsuccessful attempt.
15 ideas in, 15 out? Not quite
Perhaps the most striking figure in the report is the attrition rate of AI ideas inside Indian organisations. Only about 15 out of every 100 AI ideas proposed actually reach production. The rest die in committee rooms long before they touch a customer.
Internal technology assessments and security reviews are the two biggest executioners. Roughly 35 per cent of AI proposals do not survive IT evaluation, and another 17 per cent are rejected during security reviews. That rejection rate nearly doubles in highly regulated industries such as banking and insurance, where compliance teams hold the final veto.
Adobe calls the stuck majority the “Frozen Middle”: organisations that are piloting or even deploying AI, but without the coordination needed to scale it across the business. For agentic AI, the newer breed of autonomous systems companies are now experimenting with, 40 per cent of organisations sit in this frozen zone.
The ownership problem
What separates the 6.7 per cent, roughly 30 organisations in the survey, that Adobe classifies as AI leaders from everyone else? The report points first to ownership. Only 18 per cent of Indian companies have an AI leader with real budget authority. Among the leaders, that figure jumps to 93 per cent.
- Budget authority: Leaders give one person the purse strings for AI. Most firms do not.
- Production velocity: Leaders convert pilots to production at nearly four times the rate of other companies.
- Measurement discipline: Only 10 per cent of all organisations measure AI return on investment comprehensively. The rest track usage, a vanity metric that tells little about whether AI is actually paying off.
- Skills shortage: 82 per cent of businesses say their AI skills shortfall stayed the same or got worse over the past year.
“As AI investment grows, the gap between experimentation and enterprise-wide value is becoming harder to ignore,” said Jeremy Willmott, Senior Director of Digital Strategy Group for Asia Pacific at Adobe. He added that companies which establish clear ownership, shared measures of success and stronger collaboration between business and technology teams will be best placed to move beyond isolated pilots and achieve sustainable impact.
Why the bottleneck is structural, not technical
The report’s most important takeaway may be that the AI stall is an organisational problem, not a technology one. Most companies focus their AI scorecards on usage, how many people opened the tool, rather than indicators such as output quality, rework rates and the cost of reviewing AI-generated content. When nobody measures whether the output is actually good, pilots drift along without ever proving their worth.
Security reviews compound the problem. With 17 per cent of proposals rejected at this stage, and nearly twice that in regulated sectors, security teams have become the quiet gatekeepers of India’s enterprise AI ambitions. The report does not argue that this scrutiny is wrong, but notes that when security review comes at the very end of a project cycle, it kills months of work in a single meeting.
What this means for India Inc
The timing matters. India is positioning itself as a global AI powerhouse, with the government pushing digital public infrastructure, domestic chip ambitions and a national AI mission. Enterprise adoption is the missing half of that story. If Indian companies cannot scale AI inside their own operations, the gap between the country’s AI rhetoric and its AI reality will keep widening.
There is a silver lining. The 6.7 per cent of leaders prove that scaling is possible with today’s technology. Their playbook is not exotic: one accountable owner with a budget, security involved early instead of at the end, honest measurement of output quality, and a plan to close the skills gap rather than hoping it fixes itself.
For the 59 per cent in the frozen middle, the report’s message is blunt. Spending more without fixing ownership and measurement will just buy more expensive pilots. The Indian AI story of 2027 will be written not by who spends the most, but by who figures out how to finish what they start.
Frequently asked questions
What is the Adobe 2026 State of AI Readiness Report?
It is Adobe’s study of enterprise AI adoption, based on a survey of 475 senior executives in India. The India findings were released on October 14, 2026, and examine what separates organisations that scale AI from those that stall.
How many Indian companies have scaled AI across their business?
Only 6.7 per cent, or about 30 organisations in the survey, have scaled AI across multiple workflows and built the operating model to sustain it.
Why do most AI projects fail to reach production in India?
About 35 per cent of proposals fail internal IT evaluation and 17 per cent are rejected in security reviews. Weak ownership, with only 18 per cent of firms having an AI leader with budget authority, and poor ROI measurement are the other major blockers.
What is the “Frozen Middle” mentioned in the report?
It is Adobe’s term for the 59 per cent of organisations that are piloting or deploying AI but lack the coordination to scale it across the business. For agentic AI specifically, 40 per cent sit in this zone.
Are Indian companies spending enough on AI?
77 per cent are increasing AI investment year on year, yet nearly two in three believe they are still underinvesting. The report argues the constraint is not money but ownership, measurement and skills.
Informeia will keep tracking how Indian companies move from AI pilots to production scale. Stay tuned for the next update in this space.
