Bain & Co’s Global Technology Report 2026 says the AI industry needs $6 trillion in annual revenue by 2031 to pay for its planned infrastructure buildout. Existing consumer and enterprise applications may generate only $1.2 trillion to $1.8 trillion by then, leaving vendors to find up to $4.8 trillion in new revenue. Statista and MarketsandMarkets estimate the current AI software market at about $600 billion, implying revenue would need to grow tenfold in five years. The report compares the $6 trillion target with roughly six times the value of today’s cloud computing sector.

The article says OpenAI, Anthropic and Meta are releasing apps, assistants and platforms as the industry seeks more spending. It also cites an FT report from September that US hyperscalers issued up to $300 billion in residual value guarantees in less than a year while recording little of that exposure on their balance sheets. The guarantees promise a minimum future value for chips and data centres; the article says a shift toward smaller models running on phones and laptops would make that a high-risk bet.

Bain reports that hardware and semiconductor stocks grew at a 24% compound annual rate from 2020 to 2026, compared with 6% for software. NVIDIA is valued at $5.5 trillion, roughly four times Bitcoin’s market capitalization. On productivity, the report says developers complete about 21% more tasks with AI-powered coding, while review time increases by approximately 91%. The article connects this to the Productivity Paradox and Jevons Paradox: efficiency gains can be offset by additional work and costs.

Separate PwC research finds that 45% of UK AI users say their role has become more complex and 44% report a heavier workload. Bain estimates organizations’ IT costs could rise 75% over the same period vendors need to find $4.8 trillion in new revenue. The article says some organizations now exhaust annual AI budgets in a single quarter, despite adopting AI partly to cut costs and raise productivity.

Danielle Burgs Escobar, a Bain partner and head of its UK Enterprise Technology practice, says falling token costs can lead to more usage and higher total spending as organizations add use cases and workloads. She says consumption also grows through retries, added context and multi-step reasoning, while companies may choose newer frontier models even when simpler ones could handle a task. Escobar says organizations should measure the full cost of completing a useful task and compare it with a baseline for the former human-performed task, including its time and accuracy.

Culture Amp surveyed 112,000 workers across 123 organizations worldwide. It found that 85% said their organization actively encourages AI experimentation, while 42% did not know why they were being asked to use AI.

Bain & Co’s Global Technology Report 2026 says the AI industry needs $6 trillion in annual revenue by 2031 to pay for its planned infrastructure buildout. Existing consumer and enterprise applications may generate only $1.2 trillion to $1.8 trillion by then, leaving vendors to find up to $4.8 trillion in new revenue. Statista and MarketsandMarkets estimate the current AI software market at about $600 billion, implying revenue would need to grow tenfold in five years. The report compares the $6 trillion target with roughly six times the value of today’s cloud computing sector.

The article says OpenAI, Anthropic and Meta are releasing apps, assistants and platforms as the industry seeks more spending. It also cites an FT report from September that US hyperscalers issued up to $300 billion in residual value guarantees in less than a year while recording little of that exposure on their balance sheets. The guarantees promise a minimum future value for chips and data centres; the article says a shift toward smaller models running on phones and laptops would make that a high-risk bet.

Bain reports that hardware and semiconductor stocks grew at a 24% compound annual rate from 2020 to 2026, compared with 6% for software. NVIDIA is valued at $5.5 trillion, roughly four times Bitcoin’s market capitalization. On productivity, the report says developers complete about 21% more tasks with AI-powered coding, while review time increases by approximately 91%. The article connects this to the Productivity Paradox and Jevons Paradox: efficiency gains can be offset by additional work and costs.

Separate PwC research finds that 45% of UK AI users say their role has become more complex and 44% report a heavier workload. Bain estimates organizations’ IT costs could rise 75% over the same period vendors need to find $4.8 trillion in new revenue. The article says some organizations now exhaust annual AI budgets in a single quarter, despite adopting AI partly to cut costs and raise productivity.

Danielle Burgs Escobar, a Bain partner and head of its UK Enterprise Technology practice, says falling token costs can lead to more usage and higher total spending as organizations add use cases and workloads. She says consumption also grows through retries, added context and multi-step reasoning, while companies may choose newer frontier models even when simpler ones could handle a task. Escobar says organizations should measure the full cost of completing a useful task and compare it with a baseline for the former human-performed task, including its time and accuracy.

Culture Amp surveyed 112,000 workers across 123 organizations worldwide. It found that 85% said their organization actively encourages AI experimentation, while 42% did not know why they were being asked to use AI.