Published on the 09/04/2025 | Written by Heather Wright
As OpenAI clocks US$40b raise…
Businesses and investors are pouring billions into AI, but in the continuation of an ongoing theme many local companies are still struggling to justify the ROI of those investments.
OpenAI announced this month that it has raised US$40 billion (NZ$70b/AU$64b) in a funding round that values the maker of ChatGPT at US$300 billion.
“Generative AI is leading to a classic catch-22.”
It’s the largest venture capital deal ever and makes OpenAI the second most valuable VC-backed company, behind SpaceX, with its $350 billion valuation.
The funding is coming from Japanese tech investment group SoftBank, which will stump up $10 billion this month, followed by $30 billion in December, contingent on OpenAI transitioning to for-profit by the end of the year.
SoftBank says $10 billion of that funding will be syndicated out to unnamed co-investors.
It says it is putting up the cash because it believes OpenAI is the closest to achieving artificial general intelligence – a key milestone on the path to artificial super intelligence, which SoftBank has declared its mission ‘for the advancement of humanity’.
Massive computing power will also be essential, with SoftBank also part of the US$500 billion Stargate Project announced earlier this year amid much Trump-driven fanfare. The four year project, which also has Oracle among its investors, aims to build dedicated AI infrastructure in the United States.
VC, private equity and mergers and acquisition database Pitchbook says the OpenAI deal lifted AI deals to account for 58 percent of the total capital raised in the first quarter. Of eight deals of greater than US$1 billion, just one – Binance – was not AI focused. The ongoing AI investment comes at a time when deal numbers for non-AI companies has slumped to its lowest in a decade, Pitchbook says. The amount invested in non-AI companies ‘nearly’ surpassed the seven AI investments of $1 billion or more.
While the AI investment gold rush continues – for some at least – organisations continue to report struggles to find a return on their investment.
Cisco’s AI Readiness Index results for New Zealand show companies are committing significant investment to AI, with 97 percent of organisations feeling increased urgency to deploy AI, driven largely by the c-suite.
The report shows 42 percent of Kiwi organisations surveyed reported committing 10-30 percent of their IT budget for AI deployments.
But the report notes that despite significant AI investments in areas including cybersecurity, IT infrastructure and data analytics and management, many haven’t achieved their expected returns.
The Australian version of the report late last year found similar investment – and similar issues – with 38 percent of APJC respondents allocating 10-30 percent of their IT budget to AI and 92 percent of Australian organisations reporting increased urgency to deploy AI in the past year, with the CEO and leadership team leading the charge in creating that sense of urgency.
Other reports have suggested many AI projects aren’t progressing beyond proof of concept stage. IDC says 88 percent of AI projects aren’t making it to widescale deployment, while S&P Global says 42 percent of companies abandon most of their AI initiatives before reaching production.
Not everyone is so pessimistic, however, with the AI in Action report from New Zealand’s AI Forum, showing 56 percent of respondents reported a positive impact on financial performance and 71 percent reported savings on operating costs, with financial gains of between $1,001 and $50,000 for both growth and cost savings for ‘many’ companies.
Worker productivity is, according to respondents, a big beneficiary – some 93 percent say they’ve seen improved productivity with automation of repetitive tasks freeing employees up to focus on higher-value, strategic work, driving overall business performance.
But the accompanying report also notes an increase in companies choosing not to disclose financial impacts – possibly indicating growing sensitivity around AI-related financial data.
The survey of 200 companies shows 82 percent now report some level of AI use in their organisation, up 15 percentage points on the previous survey, six months ago.
One factor that may be aiding the ROI equation for the companies surveyed by the AI Forum is the use of existing AI apps and platforms, with 72 percent of Kiwi companies using off-the-shelf AI tools, such as ChatGPT, Microsoft Copilot and Google Gemini. Just 13 percent opting for custom-build solutions.
Unlike custom AI solutions, which can require significant investment, pre-existing AI tools typically provide a more affordable entry point and eliminate the need for costly development and specialised expertise but are less likely to meet the specific, unique, business challenges.
The use cases for those companies deploying AI successfully according to the AI Forum report also provide some insight into ROI successes.
Marketing is the top use case, followed by administration, software development, project management and design.
The AI Forum notes these are areas where companies can see an immediate return on investment, compared to industries like manufacturing, where AI integration requires significant upfront investment, planning and training.
They’re also areas which, unlike human resources or finance, typically involve less sensitive personal data, reducing privacy and compliance risks and providing less regulatory constraints.
Despite that, and lower cost options increasingly becoming available along with off-the shelf and subscription offerings, there are inevitable setup and ongoing expenses. Notably, fewer businesses reported setup or ongoing costs exceeding $50,000 compared to the previous survey.
But AI in Action also notes that the correlation between AI costs and positive business impact has weakened in the past six months.
“This shift indicates that as AI adoption matures, returns on investment are becoming more complex and dependent on business-specific factors.”
Last year Accenture warned about the dangers of introducing new technical debt through generative AI, saying the rapid deployment of the technology and other emerging tech, was leading to a surge in new technical debt.
Generative AI and AI are the highest contributors to tech debt, along with enterprise applications, according to Accenture’s digital core research.
“Generative AI is leading to a classic catch-22,” Accenture says.
“On one hand, it has the potential to create new technical debt. On the other, when used appropriately, generative AI can help manage tech debt remediation as well as minimise tech debt creation.”
