The GenAI Euphoria

The rising euphoria created by the mainstreaming of LLM via ChatGPT’s 2022 launch is often tempered by drawing parallels to a bubble. And its catch phrase – business of work will be transformed by genAI globally – appears to be settling for a more nuanced understanding.

No doubt there is discomfort when companies announce major initiatives and product plans that are not supported by available capital. OpenAI, which has already made around USD 1 trillion in AI deals¹, including a USD 500 billion data center buildout project, despite being set to generate only USD 13 billion in revenue. There is a competition among retail investors to get-in on the start-up action. AI garnered nearly half of all funding in 2025². According to JP Morgan Asset Management, AI-related stocks have accounted for 75% of S&P 500 returns, 80% of earnings growth and 90% of capital spending growth since ChatGPT launched in November 2022³.

Jeff Bezos called AI is currently in an “industrial bubble” but the technology is “real”, with inflated stock prices or valuations of companies that have disconnected from the fundamentals of a business⁴. Sam Altman ruminated about the AI bubble with reporters over a meal, calling out overexcitement about AI⁵. The tech stocks recorded a selloff on Nasdaq the next day. These concerns around AI valuations and investment sustainability have sparked debates about whether we’re witnessing another tech bubble.

Circular Financing and the Investment-Revenue Gap

Much of the reporting has focused on circular financing loops, reminiscent of creative financing in previous speculative cycles. NVIDIA’s up to USD 100 billion investment in OpenAI comes with the latter purchasing the former’s GPUs. OpenAI gets cash for its needs, NVIDIA gets future sales, and both see a bump in valuation. OpenAI to buy cloud services from Oracle, which invests in NVIDIA chips for its DCs. Both Oracle and NVIDIA see a bump in share prices; Stargate valuation rises. OpenAI and AMD. OpenAI and Microsoft. OpenAI and CoreWeave.

And yet, beyond the creative financing, the disconnect owes to the massive investment required vs the inadequate revenue levels that can justify this investment. While consumer adoption has sky-rocketed, paying ones are small and enterprises are taking a more cautious approach. There is also a growing chorus voicing the limitation of LLMs, with emerging neurosymbolic AI positioned as the answer. With the hindsight of three years and more data, there are sobering arguments supporting a rather slow, niche impact of genAI. Despite the rapid pace of generative AI adoption in consumer markets, enterprise implementation tells a different story.

Impact has been range bound to specific industries vs initial claims on multi-sector disruption driven by genAI.

Reports in 2022 invariably highlighted the transformational impact of genAI across banking and financial services (credit underwriting, risk, customer service), high tech and software (code generation, testing, DevOps, product design), healthcare (diagnostics support, documentation, triage), retail and CPG (marketing, demand forecasting, customer support), automotive, aerospace, and manufacturing. It appeared that no sector, except those that were inherently labour intensive, would be spared from its march.

Yet, the impact has been more pronounced in some areas, and less broad-based as forecasted. Claims of large scale job displacement also remain contested at best.

• Creative production has seen the most visible and discussed impact, with marketers substantially adopting it for ads, image, video, and copy. What used to be the domain of Photoshop (an Adobe product) has been widely democratized. Often cited, Klarna, a Swedish fintech, cited a USD 6 million reduction in image production costs⁶. Creative content suites have pressured content-focussed ad spend though marketers continue to reach out to them for content planning, paid media, SEO, and AI content distribution.

• The IT sector has been pushed to transform in face of genAI enabled automation. Reports from Capgemini and MIT record productivity improvement in the software engineering across tasks such as coding assistance, creating documentation, and UX design assistance. But saving at the individual level is not the same as cost saving at the organization level, for which there are few estimates. And even though deployment cases trail beyond this, tech players are betting on its ability to revamp the SDLC and help it pivot from traditional, human-centric, linear business model to one that is platform and IP driven. Most companies are reporting ongoing work on genAI agents to automate client workflows.

• Customer support, particularly chatbots, was one of the first use cases to go into production across sectors. Contact centers are benefiting from context-aware, real time suggestions and quicker resolution, leading to better customer experience. Other deployments include for knowledge management, where companies have closed models that retrieve information and develop some level of insights from internal repositories.

While companies across sectors have specific use cases, any meaningful sector-specific deployment trails beyond this. According to an MIT report of 2025, only two of eight major sectors show meaningful structural change from genAI⁷.

ROI on genAI Investments Remains Elusive

Menlo Ventures estimated 2024 business spending on generative AI at USD 13.8bn in the US alone⁸. The MIT report estimates USD 30-40 billion in enterprise investment into genAI, with little RoI to show as outcomes. Based on its sample size and data, it reports that near all, i.e. 95% of organizations are getting zero return. While they may have benefits like productivity, the same has no measurable impact on P&L.

And this reflected across surveys. A Deloitte 2025 survey across Europe and the Middle East found an unclear RoI associated with generative AI spending⁹. Informatica’s global CDO Insights 2025 survey found that >97% organizations experience difficulties demonstrating genAI’s business value¹⁰. But that hasn’t dampened the mood or spending. Its survey found that 93% of U.S. data leaders are planning to ramp up genAI investments, closely followed by APJ (86%) and Europe (82%).

The continued interest and spending on generative AI may also be partly driven by businesses wanting to ride the wave rather than waiting it out. An IBM C-suite study across 33 countries found 64% of CEOs surveyed acknowledge that the risk of falling behind drives investment in some technologies before they have a clear understanding of the value they bring to the organization¹¹.

Enterprise Grade AI Tools Are Beginning to Appear but Their Uptake and Satisfaction Remains Unknown

In insurance, Verisk has launched a Commercial GenAI Underwriting Assistant, built on Amazon Bedrock, that automates the task of insurance policy document comparison and change detection. Munich Re integrated a generative AI-driven CoPilot function into its REALYTIX ZERO underwriting platform, enabling its insurer clients to generate detailed insurance product proposals.

For law firms, genAI may seem to be the perfect solution for document review and summarization, legal research, memo and contract drafting, and knowledge management among others. Yet, a Skills.law 2025 survey¹² of over 100 of the largest firms in the US and UK had a sobering insight – Contract Companion (a proofreading tool), Intapp Time (time entry software), and Kira (contract analysis), all of which predate GPT, are the predominant tools used. Though, many of the firms have a genAI tool live, with focus on internal efficiency and productivity gains.

On the tax services side, Blue J is making a mark with a genAI tax research (using OpenAI’s GPT models), analyzing tax legislation and case law. It counts KPMG UK and RSM US as clients.

Within pharma, genAI is in prototype stage for drug discovery, with Insilico, Exscientia, and Isomorphic Labs working on small molecule generation. Others like Absci and Generate:Biomedicines are focused on biologic drug design. One of the drugs, Rentosertib from Insilico Medicine, has completed phase IIa trial and is preparing for phase IIb.

Are There Any Winners in genAI?

With genAI based solutions still trying to figure their way, monetization, and scalability, some market participants have built a robust business. While AI investment pours into the sector and AI valuations soar, a select few companies have found sustainable revenue models.

• Microchips and processors companies: NVIDIA is widely regarded to control 90% of this market, generated USD 51 billion (data center sales; 89% of total USD 57 billion) revenues in 3Q of 2025 and has set a goal of USD 500 billion in AI GPU sales by end of 2026¹³. This is part where the training of LLM happens. But here too, the market is shifting to AI inference tech, which is expected to be a much bigger market then for LLM training chips. Lending heft to this, NVIDIA made its largest purchase ever – at USD 20 billion for Groq for its low-latency inference tech¹⁴.

• Cloud infra providers: According to Synergy Research Group, Q3 2025 spending on cloud infra services was up 28% YoY, highest growth seen in the last three years¹⁵. It attributes this largely to genAI and AWS, Azure, Google Cloud, and OCI all stand to benefit from this push. It’s another note on Neoclouds, which offer GPUaaS, genAI platform services and high-capacity data centers, estimates a 205% YoY growth in Q2 2025 to USD 5 billion and forecasts to reach USD 180 billion by 2030¹⁶.

• Consulting firms: Accenture has been cashing the most on the genAI boom, with booking scaling from USD 300 million in 2023 to USD 5.9 billion for fiscal 2025, with the market estimating this to hit USD 9.3 billion in 2026¹⁷. Its genAI revenues came in at USD 2.7 billion.

• IT services companies: Though one would have expected tech MSPs to lead, the Indian ones have trailed Accenture. TCS reported USD 1.5 billion in broader AI revenues, including that from genAI¹⁸. HCLTech reported advanced AI, including genAI, revenues of USD 100 million. This is not just genAI though and falls way short of Accenture’s lead.

• LLM providers: OpenAI had USD 4.3 billion revenue during the first half of 2025 while still posting a net loss of USD 13.5 billion^{19, 20}. Its cash burn is expected at USD 115 billion through 2029. Of its 800 million users, just 5% pay. Menlo Ventures came to a similar estimate based on a USD 10 billion annual run rate and about 800 million monthly active users – about 40 million would be paying USD 20 per month to generate that much revenue²¹. While it projects a revenue run rate of USD 13 billion, it has USD 1.4 trillion in spending commitments – leaving a whole wide revenue gap and little answers. Same for Perplexity²² (USD 100 million annualized revenue in early 2025). Both Gemini and Copilot rely on subscription and report multi-million active users driven by integration of LLMs across products but shy away from breaking out paid subscribers.

Will LLMs Make Enough Money to Justify the Astronomical Investments Being Made? Or Are There Some Solutions That Are a Lighthouse?

• Coding and software engineering: Anthropic is widely believed to be ahead in genAI-driven coding, which is also its largest use-case for Claude.ai. By far, Anthropic appears to be on a better footing, with an ARR at USD 5 billion²³ by August 2025, built on an enterprise focus from API calls billed at per-token pricing. And this is expected to swell to USD 70 billion²⁴ by 2028. Github was reported to touch USD 2 billion ARR revenue in 2024, with GitHub Copilot driving 40% of the revenue growth²⁵. Cursor just hit USD 1 billion in ARR revenue²⁶.

• Content, image, video: Midjourney is one the way to USD 500 million in ARR revenue²⁷ and has often been quoted for product fitment in the consumer space. Adobe’s genAI portfolio across AI assistant, Firefly, and GenStudio clocked USD 125 million Q1 FY25, which it expected to double by end 2025²⁸. Runway expects to reach USD 300 million in 2025²⁹.

• Vertical solutions that address work flow issues: Healthcare stands out where genAI based scribing is heating up. Microsoft/Nuance’s DAX Copilot is on the top, with 1,000 providers stated to be using it³⁰. Abridge, another scribing solution, is on the way to over USD 100 million in ARR³¹. Beyond this, there are multiple verticals including professional services where genAI solutions are being deployed for task completion and research, generally resulting in agility and productivity gains.

The landscape of LLM providers and genAI enterprise solutions continues to evolve rapidly, with market dynamics shifting as companies race to prove sustainable business models.

Endnote

While the individual benefit of genAI tools remains firmly established, converting that into scalable enterprise-wise benefit, either as cost saving or revenue growth, remains at large, with few niches performing better.

Meta’s CFO Susan Li encapsulated this thinking with her comment in the Q2 2025 results conference call – “On the GenAI side, we are clearly much, much earlier on the return curve and we don’t expect that the GenAI work is going to be a meaningful driver of revenue this year or next year.”

Organizations navigating generative AI adoption and working to improve enterprise AI ROI need structured market insights into market trends, competitive moves, and proven use cases. Market intelligence and strategic research can help separate genuine opportunities from hype, identify where real enterprise value is being created, and guide disciplined investment decisions aligned with long-term business objectives.

Eliminate guesswork. Embed insight. Indigrowth.

1 A guide to the $1 trillion-worth of AI deals between OpenAI, Nvidia and others, CNBC,
^{https://www.cnbc.com/2025/10/15/a-guide-to-1-trillion-worth-of-ai-deals-between-openai-nvidia.html}
2 6 Charts That Show The Big AI Funding Trends Of 2025, Crunchbase,
^{https://news.crunchbase.com/ai/big-funding-trends-charts-eoy-2025/}
3 The Blob: the AI and data center takeover, J.P. Morgan,
^{https://am.jpmorgan.com/content/dam/jpm-am-aem/global/en/insights/eye-on-the-market/the-blob-amv.pdf}
4 Jeff Bezos hails AI boom as ‘good’ kind of bubble, FT,
^{https://www.ft.com/content/8d80f601-4725-489a-8d6e-39099a01f9cb}
5 Sam Altman says ‘yes,’ AI is in a bubble, The Verge,
^{https://www.theverge.com/ai-artificial-intelligence/759965/sam-altman-openai-ai-bubble-interview}
6 AI helps Klarna cut marketing agency spend by 25% and run more campaigns, Klarna,
^{https://www.klarna.com/international/press/ai-helps-klarna-cut-marketing-agency-spend-by-25-and-run-more-campaigns/}
7 The GenAI Divide, STATE OF AI IN BUSINESS 2025, MIT NANDA,
^{https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf}
8 2024: The State of Generative AI in the Enterprise, Menlo Ventures,
^{https://menlovc.com/2024-the-state-of-generative-ai-in-the-enterprise/#765bf53e-c1df-477f-941d-810743936402}
9 AI ROI: The paradox of rising investment and elusive returns, Deloitte,
^{https://www.deloitte.com/uk/en/issues/generative-ai/ai-roi-the-paradox-of-rising-investment-and-elusive-returns.html}
10 Global Data Leaders Seek to Harness the Power of GenAI for AI-Driven Success, Informatica,
^{https://www.informatica.com/about-us/news/news-releases/2025/01/20250128-global-data-leaders-seek-to-harness-the-power-of-genai-for-ai-driven-success.html}
11 IBM Study: CEOs Double Down on AI While Navigating Enterprise Hurdles, IBM,
^{https://newsroom.ibm.com/2025-05-06-ibm-study-ceos-double-down-on-ai-while-navigating-enterprise-hurdles}
12 SKILLS 2025 AI SURVEY RESULTS,
^{https://skills.law/skills-2025-ai-report}
13 NVIDIA Corporation (NVDA) Q3 FY2026 earnings call transcript,
^{https://finance.yahoo.com/quote/NVDA/earnings/NVDA-Q3-2026-earnings_call-379484.html}
14 Nvidia buying AI chip startup Groq’s assets for about $20 billion in its largest deal on record, CNBC,
^{https://www.cnbc.com/2025/12/24/nvidia-buying-ai-chip-startup-groq-for-about-20-billion-biggest-deal.html}
15 Cloud Market Growth Rate Rises Again in Q3; Biggest Ever Sequential Increase,
^{https://www.srgresearch.com/articles/cloud-market-growth-rate-rises-again-in-q3-biggest-ever-sequential-increase}
16 Neoclouds Currently Growing by Over 200% per Year,
^{https://www.srgresearch.com/articles/neoclouds-currently-growing-by-over-200-per-year-will-reach-180-billion-in-revenues-by-2030}
17 Accenture earnings preview: AI momentum to continue driving Q1 growth,
^{https://www.spglobal.com/market-intelligence/en/news-insights/research/2025/12/accenture-earnings-preview-ai-momentum-to-continue-driving-
q1-growth}
18 TCS Clocks USD 1.5 Bn Annual AI Revenue,
^{https://www.businessworld.in/article/tcs-clocks-usd-1-5-bn-annual-ai-revenue-583995}
19 OpenAI’s First Half Results: $4.3 Billion in Sales, $2.5 Billion Cash Burn,
^{https://www.theinformation.com/articles/openais-first-half-results-4-3-billion-sales-2-5-billion-cash-burn}
20 OpenAI’s era-defining money furnace,
^{https://www.ft.com/content/908dc05b-5fcd-456a-88a3-eba1f77d3ffd}
21 2025: The State of Consumer AI,
^{https://menlovc.com/perspective/2025-the-state-of-consumer-ai/#3d785b95-a001-440c-bd90-0d870915ddb9-link}
22
^{https://www.linkedin.com/posts/aravind-srinivas-16051987_perplexity-has-crossed-100m-in-annualized-activity-7310678232079495170-38ik/}
23 Anthropic raises $13B Series F at $183B post-money valuation,
^{https://www.anthropic.com/news/anthropic-raises-series-f-at-usd183b-post-money-valuation}
24 Anthropic Projects $70 Billion in Revenue, $17 Billion in Cash Flow in 2028,
^{https://www.theinformation.com/articles/anthropic-projects-70-billion-revenue-17-billion-cash-flow-2028}
25 Microsoft Q4 2024: CEO Nadella Calls Copilot Growth Rate Fastest For Any M365 Suite,
^{https://www.crn.com/news/ai/2024/microsoft-q4-2024-ceo-nadella-calls-copilot-growth-rate-fastest-for-any-m365-suite}
26 Past, Present, and Future,
^{https://cursor.com/blog/series-d27 https://leanaileaderboard.com/}
28 Adobe Q1FY25 Earnings Call Script,
^{https://www.adobe.com/cc-shared/assets/investor-relations/pdfs/21305202/c56hryhwgerfaw.pdf}
29 How Runway Hopes to Outrun OpenAI, Google in the AI Video Race,
^{https://www.theinformation.com/articles/how-runway-hopes-to-outrun-openai-google-in-the-ai-video-race}
30
^{https://www.linkedin.com/posts/karihackett_novant-health-is-celebrating-one-million-activity-7386385902643720193-YhvC/}
31 Abridge in Talks to Raise at $5 Billion Valuation as AI Health Startups Draw Investors,
^{https://www.theinformation.com/articles/abridge-talks-raise-5-billion-valuation-ai-health-startups-draw-investors}