Daily Update, 29 September 2026: Bain says AI may need a market approaching $6 trillion a year to fund its build-out

Bain & Company says the AI market may need to approach $6 trillion a year to pay for the infrastructure being built for it, in its Technology Report 2026, published on 29 September 2026. Overnight, Samsung and KKR announced a $1 billion Samsung commitment to Helix Digital Infrastructure, a company KKR formed to build data centres and power for hyperscalers.

Bain's arithmetic

Bain's lead article, "New Innovation Is Required to Fund AI's $6 Trillion Buildout", sets out how it reaches the figure:

• Spending. Capital expenditure by the five hyperscalers it names (Microsoft, Google, Amazon, Meta and Oracle) "could reach $780 billion in 2026", which Bain says is nearly five times the level of three years earlier. By 2031, it says, annual spending on AI infrastructure "could reach $1.5 trillion", including new data centres, compute, and upgrades to installed chips, memory and networking.
• The assumption. Bain assumes capital expenditure amounts to about 25% of industry revenue, which it calls "an ambitious but reasonable percentage based on trends among cloud providers". On that basis, it says, sustaining the spending "would require an AI market approaching $6 trillion annually".
• What is already in view. Consumer AI products "could generate an estimated $200 billion to $400 billion by 2031", and enterprise adoption "could contribute another $1 trillion to $1.4 trillion in gains to providers alone". Together, Bain says, that could total $1.2 trillion to $1.8 trillion, "leaving about $4.2 trillion of new revenue" to reach the $6 trillion it estimates is needed.

On our reading, the $6 trillion is a modelled market size resting on a stated ratio, not a forecast that the market will reach it. By our arithmetic, the $4.2 trillion gap is measured from the top of Bain's $1.2 trillion to $1.8 trillion range.

Where Bain thinks new revenue could come from

Bain lists four categories it says are "likely to help deliver this growth":

• Search and advertising: $100 billion to $200 billion "or more", from ads in chatbots and AI replacing part of traditional search.
• Autonomous everything: a $400 billion "market opportunity" across vehicles, trucks, drones and other industrial automation.
• Physical AI: a $900 billion opportunity across priority sectors, "assuming a 10% reduction in R&D and manufacturing costs".
• New product development: drug discovery, mental health support, materials science and autonomous research. Bain's text gives no figure for this category.

Bain's conclusion is that "The infrastructure is being built ahead of the demand curve", and that funding it sustainably "will require adding approximately 1% to the annual global GDP growth rate". It ends on a question: "whether the applications arrive in time to pay for it."

The build-out itself

A companion chapter, "AI Data Center Boom: Can We Build It If They Come?", gives Bain's supply-side view:

• In a base case, Bain's Data Center Model "projects approximately $5 trillion to $6.5 trillion of spending to build nearly 150 GW of new compute capacity by 2030", which Bain says would almost triple global capacity in five years.
• Bain says power, chips, skilled labour and permitting "are all constrained at once", and that the compute shortage "is unlikely to resolve on its own".
• It says local opposition "blocked or delayed at least 75 projects worth $130 billion in the first quarter of this year alone".
• It says hyperscalers' data centre capex "could top $1 trillion by 2028".

Samsung's $1 billion for Helix

KKR's release, dated New York, 28 September 2026 and carried on wire services at midnight UK time on 29 September, says Samsung "has committed $1 billion to Helix". Samsung's newsroom dates its announcement Korea, 29 September 2026.

• The split. Samsung says six of its companies are investing "a combined USD 1 billion": Samsung Electronics is investing $500 million, and Samsung C&T, Samsung SDS, Samsung SDI, Samsung Life Insurance and Samsung Fire & Marine Insurance are investing the remainder. The release does not give a figure for each of those five.
• The founding investors. KKR says Samsung's investment "builds on the more than $10 billion of capital committed to the Helix strategy at its launch by founding investors including KKR, the Kuwait Investment Authority (KIA), NVIDIA and Vistra". Neither release breaks that $10 billion down by investor.
• What Helix does. Samsung says Helix launched in June 2026 and covers hyperscale data centre development, power generation, transmission and fibre. It is led by Adam Selipsky, former chief executive of Amazon Web Services.
• Power first. Samsung's release says Helix "plans to rapidly secure energy capacity through direct investments and strategic partnerships with major energy developers, including investor Vistra".

KKR's release also notes that its investors in Helix "may have certain rights, such as priority or first look rights, to provide goods or services to Helix investments". Samsung's release says its affiliates supply chips, data centre cooling, batteries and backup power, and construction work for data centres and power plants.

Why it matters

On our reading, today's two announcements show both sides of the same bet. Bain describes a build-out that may need a market approaching $6 trillion a year to pay for itself, with about $4.2 trillion of that still to be found in markets that largely do not exist yet. Samsung and KKR show capital still being committed to that build-out, here more than $11 billion to the Helix strategy by our addition of the two releases' figures ($1 billion plus "more than $10 billion"). The two Bain chapters we read do not mention Helix, and the Samsung and KKR releases do not mention Bain, and Bain's figures are estimates resting on stated assumptions, chiefly the 25% capex-to-revenue ratio. The question Bain poses, whether the applications arrive in time, is not one today's numbers can answer.