We asked seven AI engines 59 questions about Indian business, every day for ten days, and recorded every company they named.
Seven AI engines now answer the questions investors, graduates and customers used to ask a Google search. Here is who they name, who they leave out, and where they disagree.
₹45 lakh crore
of Indian company value sits with firms that AI engines rarely name. ICICI Bank, SBI, Bajaj Finance and LIC are among them. Each is worth over ₹4 lakh crore. Each is named in under 12% of answers.
Tata Group is India's most trusted company, and it is not close.
Named in 78% of answers about corporate conduct, 14 points ahead of Infosys in second. It is the first name on 5 of the 7 engines, more than any other company manages on any subject.
On careers AI says TCS, but only just.
TCS leads on four engines with 74%. Infosys leads on the other three. Which name a job seeker hears depends on which app they open.
Reliance Industries owns the money question.
Named in 80% of answers about where to invest, and 77% of answers about who is building the future. It leads both.
India's largest carmaker appears in 2% of answers.
Maruti Suzuki builds around two in five cars sold in India. DMart, the most profitable grocer, is named in <1%. Two IT firms are named more often than India's biggest bank, insurer, refiner, carmaker and miner combined.
Ask about governance and AI names Adani. Ask about jobs and it names Infosys.
Adani is named in 61% of answers about how a company is run and 6% of answers about jobs and technology. Infosys is the mirror, 5% against 68%. One risk score would put them side by side.
Ask the same question tomorrow and half the answer changes.
Two consecutive days, same question, same engine: only 47% of the companies match. Any single search of what AI says about a company is not stable enough to conclude anything.
Two names dominate almost every answer. Behind them the field drops away fast, and the order has little to do with what these companies are worth.
Three names take almost all of the attention: TCS, Reliance and Infosys. The chart holds 32 of India's largest companies. Up the side is visibility: out of every 100 AI answers, how many name that company. Across the bottom is market cap. The lines are the median for these companies: 12% of answers and ₹4 lakh crore.
| ICICI Bank | 11% | 10.1 | |
| SBI | 7% | 9.7 | |
| Bajaj Finance | 6% | 6.8 | |
| LIC | 2% | 5.2 | |
| Sun Pharma | 6% | 4.5 | |
| Titan | 12% | 4.5 | |
| Maruti | 2% | 4.3 |
| Reliance | 64% | 17.9 | |
| Airtel | 13% | 12.1 | |
| Jio | 32% | 11.5 | |
| HDFC Bank | 22% | 11.1 | |
| Reliance Retail | 13% | 8.3 | |
| TCS | 65% | 8.2 | |
| L&T | 25% | 5.6 | |
| HUL | 16% | 4.8 | |
| Infosys | 55% | 4.5 |
| M&M | 23% | 4.2 | |
| HCLTech | 14% | 3.5 | |
| ITC | 13% | 3.4 | |
| Tata Motors | 28% | 2.9 | |
| Tata Steel | 25% | 2.3 | |
| Wipro | 22% | 1.8 | |
| TechM | 13% | 1.6 |
| Flipkart | 7% | 3.7 | |
| UltraTech | 6% | 3.4 | |
| DMart | <1% | 2.6 | |
| Asian Paints | 7% | 2.5 | |
| Coal India | <1% | 2.5 | |
| Adani Green | 5% | 2.2 | |
| Indian Oil | 2% | 1.9 | |
| Tata Power | 9% | 1.2 | |
| Tata Consumer | 6% | 1.1 |
Size does not buy a mention. Across these 32 companies market cap and visibility are only weakly linked.1
The bottom right corner holds the companies most Indians actually deal with. ICICI Bank, SBI, Bajaj Finance, LIC, Sun Pharma, Titan and Maruti: between them the bank accounts, the life policies, the loans and the cars of a very large part of the country. All 7 are worth over ₹4 lakh crore. All 7 are named in under 12% of answers. Someone asking an engine about their own bank is being handed a different bank's name.
Two banks of the same size, one named twice as often. HDFC Bank and ICICI Bank are within 9% of each other in value. HDFC Bank is named in 22% of answers, ICICI Bank in 11%. Their size does not explain the difference. Being named is a separate thing from being big, and it is earned separately.
The gap runs both ways. M&M, HCLTech, ITC and Tata Motors are each named more often than a company worth at least 2.8 times as much.
The market and the engines are answering different questions. Only one of those answers is what a job seeker, a customer or a first-time investor now reads.
The widest gaps between market cap and visibility
Rank by market cap against rank by visibility, among these 32
| Company | Value rank | Named rank | Gap |
|---|---|---|---|
| Life Insurance Corporation of India | 11 | 28 | −17 |
| Bajaj Finance | 9 | 24 | −15 |
| Bharti Airtel | 2 | 16 | −14 |
| State Bank of India | 6 | 20 | −14 |
| Maruti Suzuki | 16 | 30 | −14 |
| ICICI Bank | 5 | 18 | −13 |
| Sun Pharmaceutical | 14 | 26 | −12 |
| Reliance Retail | 7 | 15 | −8 |
Life Insurance Corporation of India is 11th by value and 28th by how often it is named. That is a gap of 17 places.
companies are worth over ₹4 lakh crore and yet named in under 12% of answers. Between them, about ₹45 lakh crore of value.
A house is worth every listed company under it, so Tata Group and TCS are never plotted together. Adding them would count TCS twice.
One house is in a category of its own, and the rest are not close. Tata Group, Adani Group, Aditya Birla and Mahindra Group clear both lines: worth over ₹4 lakh crore and named in more than 12% of answers. Tata Group is named more than twice as often as any other house here. Bajaj Group and JSW Group are the reverse: large, and rarely named as houses at all. Godrej Group, at ₹2.1 lakh crore across 5 listed companies, is the only one worth under ₹4 lakh crore.
1 The correlation between market cap and visibility across these 32 companies is 0.34, on ranks rather than raw figures. 0 would mean no link at all, 1 a perfect one.
How to read it. The x-axis is market cap, on a log scale. The y-axis is visibility: the % of AI answers that name the company.
What the corners mean. The lines are the median for these 32 companies: ₹4 lakh crore and 12% of answers. A company sitting just either side of a line is a borderline case, not a clear member of that corner.
Where the market cap figures come from. Read on 18 August 2026 for the listed companies. Three are not listed, so each carries the newest valuation its owner has published: Jio (IPO-implied valuation), Reliance Retail (last disclosed valuation), Flipkart (buyback-implied valuation). Jio's comes from its own draft prospectus, filed with the regulator on 19 June 2026. Reliance Retail's is from 2023. No newer transaction has been announced, and it is the oldest number on either chart.
Jio and Reliance Retail are inside Reliance Industries. They appear separately because AI engines name them separately. Their value is not on top of Reliance Industries' own.
House totals add up the listed companies, with one exception. Where one company in a house owns part of another, that value is counted twice. This matters most for the Bajaj Group, where Bajaj Finserv holds about half of Bajaj Finance, and for the Aditya Birla Group, where Grasim holds most of UltraTech Cement. Vodafone Idea is the exception. The Aditya Birla Group lists it as one of its businesses, but the Government of India owns 49% of it. We count it at the 9.57% the group held at the June 2026 quarter. That is ₹14,651 crore, not ₹1,53,092 crore. That holding rises to 13.02% once the warrants approved in June 2026 convert in full. Correcting the remaining cross-holdings would not move any house into a different corner.
Market cap is not something this study measures. It is market data, used here only for this comparison.
The index asks the four things people actually ask about a company: who can be trusted, where to build a career, where to invest, and who is building what comes next. A different company leads each one, and the questions behind each ranking are printed alongside it.
No company leads all four subjects, and no subject gets the same answer from all seven engines.
Tata Group
named in 78% of answers, and first on 5 of the 7 engines
What was asked 11 questions
Which Indian business groups are most strongly associated with contributing to India's development beyond simply making profits?
Which major Indian business groups have created the greatest positive social impact in India, and through what work?
Tata Consultancy Services
named in 74% of answers, and first on 4 of the 7 engines
What was asked 11 questions
I'm comparing offers from large Indian companies. Which employers offer the best overall mix of compensation, career growth and work-life balance?
I'm a data or AI professional. Which large Indian companies are doing serious AI work and would be good places to build my career?
Reliance Industries
named in 80% of answers, and first on 3 of the 7 engines
What was asked 11 questions
Among India's major business groups, which have the most credible succession plans and strongest leadership continuity?
Which Indian conglomerates have the best record of allocating capital across businesses without destroying shareholder value?
Reliance Industries
named in 77% of answers, and first on 5 of the 7 engines
What was asked 11 questions
I'm a startup founder in India. Which large companies are genuinely good at partnering with startups or investing in new ventures?
Among India's large business groups, which have the strongest record of incubating a new business and scaling it successfully?
Each card shows two of the questions behind it. Every question in the index is listed in the methodology section.
Each question has a clear leader and none has a unanimous one, so which company a person meets first depends on which app they happened to open.
Trust is the nearest thing to a settled answer. The Tata Group is the first name on 5 of the 7 engines. ChatGPT names Infosys and Copilot names Reliance Industries.
Careers are genuinely contested. TCS leads on ChatGPT, Gemini, Google AI Mode and Google AI Overviews. Infosys leads on Claude, Copilot and Perplexity.
Investment splits two ways. Reliance Industries is named in more answers overall (80% against 77%), but TCS is the first name on more engines, 4 against 3. Which company an investor meets first depends on the tool.
Innovation is the most consistent. Reliance leads on 5 of the 7. It is the one question where a company can reasonably be called the AI answer.
No company leads on all four, and none of the seven engines agrees with the other six on all four. There is no such thing as being the AI answer about Indian business.
Which company each engine names most, by question
The most-named company in that engine's answers
| AI engine | Trust | Careers | Investment | Innovation |
|---|---|---|---|---|
| ChatGPT | Infosys | TCS | Reliance | Reliance |
| Gemini | Tata Group | TCS | TCS | Reliance |
| Claude | Tata Group | Infosys | Reliance | TCS |
| Copilot | Reliance | Infosys | TCS | Infosys |
| Perplexity | Tata Group | Infosys | Reliance | Reliance |
| Google AI Mode | Tata Group | TCS | TCS | Reliance |
| Google AI Overviews | Tata Group | TCS | TCS | Reliance |
Source: India AI Brand Index 2026.
A company that has checked one engine and found itself on top has checked one seventh of the picture.
Ask an AI engine about Indian business and the picture that comes back is built almost entirely from software companies. The firms that refine the fuel, insure the households, mine the coal and sell the groceries barely appear.
Market leaders, and how often AI names them
% of AI answers naming the company
Source: India AI Brand Index 2026. Every figure checked against the original answer text.
Two IT companies are named more often than India's largest bank, insurer, refiner, carmaker, miner and grocer put together.
Maruti Suzuki builds around two in every five cars sold in India. It is named in 2% of answers. DMart, the most profitable grocery chain in the country, appears in <1%. Coal India, the largest coal producer on earth, manages <1%.
It is not one engine's quirk. Maruti Suzuki is named in between <1% and 5% of answers depending on the engine: low on every one of the seven, on all ten days.
The missing names have something in common. Most sell mainly inside India, and several are state owned: Coal India, Indian Oil, LIC, the State Bank of India. They earn enormous revenue and employ vast numbers of people. They attract little English-language coverage.1
These are not obscure companies. They are the ones a graduate applies to, a household insures with and a driver buys from. On the tools those people now ask, they are close to absent.
DMart
India's most profitable grocery retailer
Coal India
The world's largest coal producer
Maruti Suzuki
About 40% of every car sold in India
1 That last sentence is a possible reason, not a finding. This index cannot establish why a company is named as often as it is.
A whole-market table flatters the biggest names. Read sector by sector and the order changes.
Source: India AI Brand Index 2026. % of answers naming the company.
Being the biggest firm in a sector does not put a company first in it. India's largest bank is third in banking. Its largest carmaker is seventh in automotive.
Banking is the clearest case. HDFC Bank leads at 22%. ICICI Bank follows at 11%. The State Bank of India, by far the country's largest bank, comes third at 7%: under half the leader.
Automotive is the furthest from the market. Tata Motors and Mahindra & Mahindra take the top two places while Maruti Suzuki, which builds around two in every five cars sold in India, sits seventh.
The market leaders AI demotes have something in common. The State Bank of India is state owned, and so is Coal India, the largest coal producer on earth, which appears in under 1% of answers. Maruti Suzuki sells almost entirely inside India, through showrooms. None of them is the kind of company written about in English every quarter.1
The spread inside a sector matters as much as the order. FMCG is a real contest, Hindustan Unilever 16% to ITC 13%. Metals is not one at all: Tata Steel is named five times as often as Hindalco, so a reader asking about Indian steel meets one company.
1 A possible reason, not a finding. This index records how often a company is named. It cannot establish why.
Some houses are famous as houses. Others are famous through the companies inside them. The Tata name alone is named in 65% of answers; the HCL name in 3%, though HCL companies reach 16%.
Some Indian houses are known as a house. Others are only known through the companies inside them, and the difference decides where a group's reputation can be built.
Tata is the rare case where the family name does the work by itself. An engine says "the Tata Group" in 65% of answers before any individual company is counted. Add TCS, Tata Motors, Tata Steel, Titan and eighteen others and some part of the group appears in 94%. Nobody has to know that TCS is a Tata company for Tata to be credited.
Aditya Birla is the opposite, and it is the more common shape. The group name reaches 24% of answers and the whole group only 25%. UltraTech, Hindalco and Grasim are each well known in their own market and pass almost nothing back up to the parent.
Neither is better, but they are not interchangeable. A house whose name carries on its own can move a new business quickly. A house of separate names has to earn each one again, and cannot borrow from a strong sibling.
Select any house to see inside it.
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Most of what AI says about Tata Group comes through its operating companies rather than the house name.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Reliance Industries is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Infosys is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Most of what AI says about Mahindra Group comes through its operating companies rather than the house name.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Larsen & Toubro is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Most of what AI says about HDFC Group comes through its operating companies rather than the house name.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
point added by its companies
companies named in 1% or more
Aditya Birla Group is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
company named in 1% or more
Only one name in Wipro is named in 1% or more of answers.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Adani Group is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
company named in 1% or more
Only one name in Hindustan Unilever is named in 1% or more of answers.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
The HCL Group name barely registers on its own. Its companies carry it.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
The Bharti Enterprises name barely registers on its own. Its companies carry it.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
company named in 1% or more
Only one name in ITC is named in 1% or more of answers.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
point added by its companies
companies named in 1% or more
Godrej Group is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
companies named in 1% or more
Bajaj Group is visible mainly as a house. Its operating companies add little on top.
Inside the house
% of answers naming the company
name it by the house name
name any part of the group
points added by its companies
company named in 1% or more
Only one name in ICICI Bank is named in 1% or more of answers.
Inside the house
% of answers naming the company
Source: India AI Brand Index 2026.
Twelve of the 59 questions ask about risk: governance concerns, poor job security, greenwashing, crises handled badly. They are reported separately and never averaged with the other 47. Read together they look like one subject. They are two, and the companies on each are almost entirely different.
Being named in an answer about risk is not the same as being accused of anything.
Six of the twelve questions ask how a company is run: governance, environment, debt, succession, group structure. The other six ask whether it can look after its people and keep up with technology: layoffs, promotions, workplace culture, AI claims, legacy systems. A company that dominates one list is almost always missing from the other.
Adani owns the first kind. It is named in 61% of answers about how a company is run and 6% of answers about jobs and technology. Infosys is the exact mirror: 5% and 68%. On one combined risk score they would sit side by side. They have almost nothing to do with each other.
That mirror explains something that otherwise looks like a contradiction. Infosys and TCS are among the most praised companies in this report, and they are also near the top of the risk list. Ask an engine which Indian companies face the hardest questions about layoffs, promotions or whether AI will take the work, and it reaches for the largest IT employers. Size is the qualification, not conduct. TCS (69%), Wipro (62%) and HCLTech (34%) are there for the same reason, and on the governance questions none of them clears 12%. Wipro is the one company here named more often in risk answers than anywhere else, and every point of that comes from the jobs and technology side, not from governance.
Which way a company leans matters more than how often it is named. Reliance appears in 42% of risk answers, more than Adani's 33%. But Reliance is named in 64% of the other 47 questions, so it comes up less once the subject turns to risk. Adani runs the other way, 18% against 33%, and that swing is the widest in the index. Vedanta is sharper still, named roughly five times as often when the question is governance.
The Tata Group leans the safest way of any name here, 31 points, and Jio, Mahindra, Larsen & Toubro and HDFC Bank lean the same way. Being named often and criticised rarely is the strongest position on this page, and Tata holds it by the widest margin in the index.
AI does not forget. Satyam, whose fraud broke in 2009 and which has not traded under that name since, is still named in 16% of answers about how a company is run. IL&FS, which defaulted in 2018, and Byju's sit in the same set. A collapse leaves the news cycle years before it leaves the answer, and no communications team gets to decide when.
The year Satyam's fraud came to light. It is still named when AI engines are asked who handled a crisis badly.
Which way each company leans
The gap between the two, in percentage points
| Company | The other 47 questions | The 12 risk questions | Gap |
|---|---|---|---|
| Adani Group | 18% | 33% | +15 |
| Vedanta Limited | 3% | 17% | +14 |
| Wipro | 22% | 32% | +10 |
| Byju's | <1% | 9% | +9 |
| Satyam Computer Services | <1% | 8% | +8 |
| Vedanta Group | 1% | 8% | +7 |
| Tata Group | 65% | 34% | -31 |
| Tata Consultancy Services | 65% | 40% | -25 |
| Jio | 32% | 9% | -23 |
| Reliance Industries | 64% | 42% | -22 |
| Mahindra Group | 32% | 10% | -22 |
Source: India AI Brand Index 2026. 834 answers to the 12 risk questions, 3,262 to the other 47.
Who AI names when the question is how a company is run
Governance, environment, debt, succession, group structure. 415 answers
Who AI names when the question is jobs and technology
Layoffs, promotions, workplace culture, AI and innovation claims, legacy systems. 419 answers
Source: India AI Brand Index 2026.
What was asked 12 questions
Which Indian conglomerates have the most complicated structures for a public-market investor to understand?
Among India's major business groups, where should investors be most concerned about succession or dependence on a few key leaders?
Which large Indian companies currently have the biggest governance, debt or promoter-related risks that I should examine carefully?
Counter-examples have been taken out. An answer to a risk question often names a company as the opposite of what was asked: as the group that handled a crisis well, as the one rated low risk in a comparison table, or in a sentence saying it should not be on the list at all. All 834 answers were read one mention at a time and 189 such mentions were removed from this chapter. They still count everywhere else in the report, because the company was still named.
What these figures do and do not say. They record how often a company is named in answers to a risk question. They do not say the engine judged the company harshly, and they are not evidence that any criticism is accurate. A company can be named as the largest firm in a sector under scrutiny. We are reporting what the engines said. We are not endorsing it.
There is no single AI answer about an Indian company. There are seven, and they do not agree on who belongs.
How each engine answers
Companies named per answer, the company it names most, and how often it credits a source
| AI engine | Companies per answer | Names most often | Share | Credits a source |
|---|---|---|---|---|
| Claude | 15.6 | Infosys | 71% | 37% |
| ChatGPT | 14.5 | Reliance Industries | 70% | 39% |
| Copilot | 12.5 | Infosys | 66% | 12% |
| Gemini | 11.7 | Tata Consultancy Services | 76% | 19% |
| Google AI Mode | 10.8 | Tata Consultancy Services | 76% | 26% |
| Google AI Overviews | 7.3 | Tata Group | 67% | 12% |
| Perplexity | 6.8 | Reliance Industries | 64% | 27% |
Source: India AI Brand Index 2026.
The blind spot is on every engine
% of that engine's answers naming Maruti Suzuki
Source: India AI Brand Index 2026.
Six engines call TCS an investment. One does not.
% of that engine's answers naming TCS
Source: India AI Brand Index 2026. 763 answers to the investment questions, 768 to the careers questions.
Each engine has a company it names most, and it names that company at about the same rate as every other engine does its own: between 64% and 76%. What changes is how much room is left behind it.
Claude names 15.6 companies in a typical answer. Perplexity names 6.8. So a company that is not one of the two or three obvious answers is competing for roughly six places on one engine and sixteen on another. For the biggest names the choice of engine barely matters. For everyone else it decides whether they appear at all.
The engine that reaches the most people shows the fewest companies. Google AI Overviews appears above ordinary Google results whether or not anyone asked for it, and it is the second most sparing of the seven at 7.3 companies per answer.
They also differ in how often they show their working. ChatGPT credits a source in 39% of its answers and Claude in 37%. Copilot and Google AI Overviews do so in 12%. The engine that reaches people without being asked for is also the one least likely to tell them where the claim came from.
On one company they do not differ by degree. They disagree. Ask six of the seven which large Indian companies to invest in and TCS is named in 75% to 94% of answers. Ask Perplexity and it is 36%.
That is not Perplexity being sparing. On careers it names TCS in 72% of answers, squarely inside the range the other six give it, and on investment it still names Reliance in 77%. Move within Perplexity from the careers questions to the investment questions and TCS falls 35 points. On five of the other six it rises.
The pattern holds on the next question along. Perplexity names TCS in 36% of answers about who is building what comes next, against 65% on the others. Investment and innovation are the two questions on which the case against Indian IT services rests, and they are the two where this engine has moved.1
How often Perplexity names TCS in answers about where to invest, against the average of the other six engines. On careers the two are three points apart.
1 Indian IT shares have been marked down over the past year on the argument that AI erodes the services model. This index does not measure share prices and cannot say the two are connected. It records that six engines describe TCS the way the market did before that argument, and one does not.
The averages in this report are steady. The individual answers behind them are not: between one day and the next, roughly half the companies in an answer change.
Anyone who has checked what AI says about their company has almost certainly checked it once. That is about half the picture.
Ask the same engine the same question on two consecutive days and only 47% of the companies match. A week apart it is 39%, and it keeps falling as the gap widens, so the picture is moving rather than wobbling around a fixed one.
This is why the figures in this report are averages over ten days and seven engines, and why no single answer is quoted as evidence of anything. A company named on nine days out of ten can be absent on the tenth.
It cuts both ways for anyone reading a screenshot. One answer that names a rival and not you proves very little. So does one that names you.
of all company positions appeared on exactly one of the ten days. Anyone checking on the other nine would never have seen them.
How much two answers to the same question share, by days apart
% of companies named in both answers
Source: India AI Brand Index 2026. 17,265 answer pairs.
How often a company held its place across all ten days
% of a company's appearances that held on nine days or more
Source: India AI Brand Index 2026. 12,656 positions observed.
Even India's most-named companies hold their ground barely half the time. Larsen & Toubro holds it once in seven.
Not the annual report. Employee review sites are credited more often than Reuters, and a proxy advisory firm outranks every global consultancy.
Sources AI names when backing up a claim about an Indian company
Number of answers crediting each source. 283 organisations were credited in total.
Source: India AI Brand Index 2026. Counts are answers, not links.
When an engine backs up a claim about a company, it names where the claim came from.
The Economic Times leads. LinkedIn, Glassdoor, AmbitionBox and Great Place to Work then take four of the next five places. Content written by employees is credited more often than the newswire.
Two specialist names sit unusually high. IiAS, the proxy advisory firm, is credited 58 times. Ethisphere, which publishes an ethics ranking, 53 times. Both outrank PwC, Deloitte, McKinsey and S&P Global.
And Hindenburg Research, one short-seller report published in 2023, is still credited 18 times as evidence about how Indian companies behave.
Not one of the 18 most-credited names is a company's own. No annual report, no press release, no corporate website. What an engine says about an Indian company is assembled almost entirely from what other people wrote about it, and a large share of that was written by its own employees.
Times employee review sites were credited, against 85 for Reuters.
No company is equally present in every conversation, and the shape of the gap is usually more useful than the overall rank. TCS has the widest spread of any major Indian company: it is the careers answer and it is not the trust answer. Green is where a company stands furthest above its own average across the four subjects, red furthest below, so a small company and a large one can be read the same way.
| Company | Strongest subject | Weakest subject |
|---|---|---|
| Tata Group | Trust +13 | Careers -8 |
| Tata Consultancy Services | Investment +13 | Trust -20 |
| Reliance Industries | Investment +18 | Careers -20 |
| Infosys | Careers +12 | Investment -15 |
| Mahindra Group | Trust +5 | Careers -4 |
| Jio | Innovation +18 | Trust -20 |
| Tata Motors | Innovation +15 | Trust -10 |
| Tata Steel | Trust +12 | Investment -6 |
| Larsen & Toubro | Investment +18 | Trust -8 |
| Aditya Birla Group | Investment +7 | Innovation -9 |
| Mahindra & Mahindra | Innovation +9 | Careers -12 |
| HDFC Bank | Investment +21 | Careers -12 |
| Wipro | Careers +12 | Investment -16 |
| Adani Group | Investment +15 | Careers -16 |
Figures are percentage points above or below that company's own average across the four subjects.
All 59, exactly as they were put to the engines on each of ten days. The wording was never changed between days or between engines.
Questions marked with a red rule ask about risk, red flags or poor conduct. Their answers appear only in the criticism chapter and are never averaged with the other 47.
Listed exactly as AI names it. Houses and the companies inside them are counted separately, never added together.
| # | Brand | % of answers | House | |
|---|---|---|---|---|
| 1 | H | Tata Group | 65% | |
| 2 | C | Tata Consultancy Services | 65% | Tata Group |
| 3 | C | Reliance Industries | 64% | |
| 4 | C | Infosys | 55% | |
| 5 | H | Mahindra Group | 32% | |
| 6 | C | Jio | 32% | Reliance Industries |
| 7 | C | Tata Motors | 28% | Tata Group |
| 8 | C | Tata Steel | 25% | Tata Group |
| 9 | H | Larsen & Toubro | 25% | |
| 10 | H | Aditya Birla Group | 24% | |
| 11 | C | Mahindra & Mahindra | 23% | Mahindra Group |
| 12 | C | HDFC Bank | 22% | HDFC Group |
| 13 | C | Wipro | 22% | |
| 14 | H | Adani Group | 18% | |
| 15 | C | Hindustan Unilever | 16% | |
| 16 | C | HCLTech | 14% | HCL Group |
| 17 | C | ITC | 13% | |
| 18 | C | Tech Mahindra | 13% | Mahindra Group |
| 19 | C | Reliance Retail | 13% | Reliance Industries |
| 20 | C | Bharti Airtel | 13% | Bharti Enterprises |
| 21 | C | Titan Company | 12% | Tata Group |
| 22 | C | Tata Sons | 12% | Tata Group |
| 23 | C | ICICI Bank | 11% | |
| 24 | H | Godrej Group | 11% | |
| 25 | C | Microsoft | 10% | |
| 26 | C | Amazon | 10% | |
| 27 | C | Accenture | 10% | |
| 28 | C | 9% | Alphabet | |
| 29 | C | Tata Power | 9% | Tata Group |
| 30 | C | State Bank of India | 7% | |
| 31 | C | Asian Paints | 7% | |
| 32 | H | Bajaj Group | 7% | |
| 33 | H | Flipkart | 7% | |
| 34 | H | JSW Group | 7% | |
| 35 | C | Tata Trusts | 7% | Tata Group |
| 36 | C | Tata Consumer Products | 6% | Tata Group |
| 37 | C | Bajaj Finance | 6% | Bajaj Group |
| 38 | C | UltraTech Cement | 6% | Aditya Birla Group |
| 39 | C | Sun Pharmaceutical | 6% | |
| 40 | C | Adani Green Energy | 5% | Adani Group |
C Company. An operating business named in its own right, such as TCS or Tata Motors. H House. An umbrella business group named as a group, such as the Tata Group or Aditya Birla Group.
A house and a company inside it are separate entries because AI engines name them separately. Different names for one company are counted once: Reliance, Reliance Industries and RIL are a single entry.
Everything in this report comes from answers seven AI engines actually gave, recorded on ten separate days. This section explains how, and what the figures can and cannot show.
59 questions, written the way a person would ask them rather than as keywords. They cover four subjects: investment, careers, innovation and trust. Three more ask which Indian companies come to mind at all.
47 ask who performs well. 12 ask about risks, red flags and poor conduct.
The two sets are never mixed. Anywhere a higher number reads as better, we count only the 3,262 answers to positive questions. That covers the subject tables, the sector tables, the business houses, the missing-companies chapter, the value chart and the reference table. The criticism chapter counts only the 834 answers to the 12 risk questions. Otherwise a company named often in answers about governance failures would score as if it had been praised.
Three measures use all 4,096 answers, because they are not about merit: day-to-day reliability, the sources engines credit, and how many companies each engine names.
Every question was put to ChatGPT, Gemini, Claude, Copilot, Perplexity, Google AI Mode and Google AI Overviews, on each of ten days between 3 and 17 August 2026.
That produced 4,096 complete answers. Asking on ten days stops one odd answer from moving a result. It is also how we measured the day-to-day change reported in chapter 8.
Every answer was read in full and every company named in it recorded. A company counts once per answer however often it appears, so a long answer cannot inflate a score by repetition.
Only companies named in the visible text count. A company appearing solely inside a web address is not a mention. Where an engine credits a source for a claim, that is recorded separately from the companies being discussed.
Different names for one company are counted once: Reliance, Reliance Industries and RIL are a single entry.
A business house and a company inside it are counted separately, because the engines name them separately. Holding companies and philanthropic bodies are left out of the company tables.
A house's total footprint is the % of answers naming any part of the group. It is not the sum of its companies. Adding those would count the same answer more than once.
Companies close together in the full table are not meaningfully apart. Several sit a fraction of a percentage point from each other. A different but equally reasonable set of questions would move them by more than that.
The sector tables and the four-subject tables are the more reliable comparisons, because they set like against like. Percentages are rounded to whole numbers. Anything under 1% is shown as under 1% rather than rounded up.
This index measures which companies AI engines name, and in answer to what. It does not measure whether an engine spoke well or badly of a company. No figure here is a verdict on any company.
Being named often in answers about risk means a company is part of that conversation, sometimes as an example and sometimes as a counter-example. It is not a finding of wrongdoing.
The index cannot explain why a company is named as often as it is. Where the report offers a reason, it says so.
Sign in and the PDF starts straight away.
The complete report as a PDF, with every sector and business house laid out in full.
ChatGPT
Gemini
Claude
Copilot
Perplexity
Google AI Mode
Google AI Overviews
