Artificial intelligence has become one of the forces holding up the British economy, but not yet in the way its promoters promised. According to the Bank of England's November 2026 Monetary Policy Report, published on November 5, 2026, novel AI-related activity accounts for "just under a fifth of GDP growth since the start of 2026." The growth comes from building and selling AI, not from using it. Firms surveyed by the Bank estimate that AI has raised their productivity by around 0.5% over three years, and in aggregate they report close to zero effect on employment.

The report, which accompanied the Monetary Policy Committee's 8-1 vote to raise Bank Rate by 0.25 percentage points to 4%, devotes most of its pages to an energy price shock from the conflict in the Middle East. Yet AI runs through it in five places: a new Bank staff proxy that tracks AI activity in the national accounts, the strength of business investment, a box on productivity, the hiring decisions of British firms, and the price of memory chips arriving from Asia. Taken together, the Bank's assessment is that the AI roll-out so far looks like a boost to demand. Whether it turns into a boost to supply, which would ease inflation rather than add to it, is the question the Committee says it will keep watching.

In Short

The Bank of England looked at how much artificial intelligence is helping the UK economy and found that companies building and selling AI services produced almost a fifth of this year's growth. Most other businesses using AI have seen only a small rise in how much each worker produces, and AI has not yet caused overall job losses, although some firms are hiring fewer people because of it. That matters for interest rates: spending on AI pushes prices up now, while AI only helps bring inflation down if it makes workers much more productive later.

The decision, briefly

At its meeting ending on November 4, 2026, the MPC voted by a majority of 8-1 to increase Bank Rate to 4%, according to the report. One member voted for a larger increase of 0.5 percentage points, to 4.25%. The rate had been held at 3.75% at the previous meeting, which ended on September 16, 2026.

The trigger was energy. "The conflict in the Middle East is now in its ninth month," the Committee wrote, and global energy prices "have continued to rise from an already elevated level." Consumer price inflation reached 3.5% in September, up from 2.6% in June, and is projected to average 4.9% in the second quarter of 2027 before falling back. The Committee found "limited evidence of knock-on effects on other prices and wages" but judged that "the degree of inflationary pressure in the pipeline is now sufficient to warrant an increase in Bank Rate."

Activity mattered too. "Activity data in recent months have generally been stronger than expected, adding to the case for modestly tighter monetary policy," the summary states. In July, the Committee had expected a weak economy to limit companies' pricing power and workers' ability to bargain for higher wages. That cushion is now thinner. And one of the reasons the economy has been stronger than expected, by the Bank's own account, is AI.

How the Bank measures AI in GDP

Official statistics do not have a line for artificial intelligence. To get around that, Bank staff have built what the report calls an "AI activity proxy" to "help monitor the effects of AI on the economy." It combines the output of three industries under the UK's standard industrial classification:

  • SIC 58, publishing activities, which the Bank notes includes developing generative AI agents and AI application software
  • SIC 62, computer programming, consultancy and related activities, which includes consulting on AI model selection and systems
  • SIC 63, information service activities, which includes AI data processing and data centre services

The Bank is open about the weaknesses. "Not all the output of these sectors will capture AI activity," the report states. The publishing division in the official classification also covers books, newspapers and periodicals, not only software, and computer consultancy predates large language models by decades. To support the link, the report cites the Office for National Statistics, which "notes that many of the businesses reporting the strongest turnover growth within these industries are involved in AI-related and cloud-computing activities."

There is a second adjustment. Some of the growth in these industries, the report says, "may also reflect the longer-run expansion of technology-related activity." Bank staff therefore strip out that trend before attributing growth to AI. What remains is the estimate behind the headline: "Adjusting for this trend suggests that novel AI-related activity accounts for just under a fifth of GDP growth since the start of 2026."

What the chart shows

Chart C.1 in the report, titled "Output growth has been resilient in the year to date, helped somewhat by AI-related activity," splits GDP growth by sector. On PPC Land's reading of the chart, the AI activity proxy contributed around 0.1 percentage points to three-month-on-three-month GDP growth in March and April 2026, rising to roughly 0.2 percentage points by July and August. The Bank publishes the underlying values in a spreadsheet alongside the report; the figures here are approximations taken from the bars, not numbers printed in the text.

The broader picture is one of strong services. GDP grew by 0.5% in the second quarter of 2026 and is expected to have grown 0.7% in the third, according to the report, against July projections of 0.3% and 0.1%. Business-to-business services, "such as information and communications services," contributed around 0.4 percentage points to quarterly growth in the second quarter. In the three months to August, GDP grew 0.7%, of which business-to-business services contributed 0.6 percentage points. Bank staff expect four-quarter GDP growth of around 2.2% in the fourth quarter of 2026, up from a little below 1.0% a year earlier.

Upstream, not across the economy

The location of the growth matters. "The contribution to growth from AI to date appears to reflect activity in the upstream development and implementation of AI technologies rather than broader gains from their adoption elsewhere in the economy," the report states. In other words, firms that build, host and install AI systems are growing. The retailers, banks, manufacturers and service businesses that buy those systems are not yet showing the gains in the aggregate data. The Bank adds that this pattern "is consistent with the impact on productivity having been relatively small so far."

Investment: the fastest two quarters since 2022

Business investment is where the AI build-out shows most clearly. According to the report, business investment grew by 3% in the first quarter of 2026 and 1.8% in the second, "the fastest growth over two quarters since 2022." Bank staff project a further rise of around 1% in the third quarter. Over the first half of 2026 alone, investment grew "by nearly 5%," although the Bank warns the data "can be volatile and subject to large revisions."

The Bank ties part of that strength to AI. "Within that, investment in the information and communication sector has also grown strongly, some of which likely reflects AI-related investment," the report states. Intelligence from the Bank's regional Agents, who gather information from businesses across the country, "also points to AI-related spending as having been a driver of recent business investment growth." Investment in utilities accounts for another part.

The revision to the forecast is large. In the central projection tables, business investment grows 4.3% in 2026, against a July projection of a 1.5% fall. For 2027, the projection rose to 2.5% from 0.9%. The Bank does not attribute the whole revision to AI, and "some moderation in business investment growth is expected in the near term following its recent strength."

The global channel

AI also reaches the UK through its trading partners. The report credits "strong AI-related investment in the US, strong AI-related exports in Asia, and increased fiscal spending on defence and infrastructure in the euro area" with keeping global growth resilient despite the energy shock. Four-quarter UK-weighted world GDP growth was 2.3% in the second quarter of 2026 and is estimated at 2.1% in the third, 0.4 and 0.3 percentage points higher than expected in July.

That US investment has been visible in company accounts covered by PPC Land. Alphabet's capital expenditure reached $51.4bn in the third quarter of 2026, leaving the company with negative free cash flow of $9.1bn despite $26.2bn of profit. S&P Global Ratings moved Meta's outlook to negative on October 30 after forecasting more than $225bn of capital spending for 2027. These are the hyperscaler budgets that, on the Bank's account, have helped sustain demand for British exports.

Productivity: a small effect, so far

Box D of the report, "Developments in UK productivity growth," is where the Bank weighs the evidence most carefully. It starts with a statistical change. In September 2026, the ONS announced a new "component approach" to measuring labour inputs, replacing estimates drawn from the Labour Force Survey, which suffers from low response rates. Under the new method, output per hour worked increased by an average annual rate of "close to 1% over 2009-24, compared to around 0.5% prior to the measurement changes." The ONS will publish quarterly estimates under the new method, up to the third quarter of 2026, on November 17.

Proxy measures suggest that output per job and output per worker have grown at an annual pace of around 1.5% recently, according to the report. That is above the MPC's assumption of 1% trend productivity growth. The question is why.

Three explanations

The report sets out three candidates in a table:

  • Cyclical adjustment: firms unwinding post-pandemic labour hoarding, with "likely no material persistent effect on productivity growth"
  • Rise in labour costs: firms reducing employment in lower-paid occupations after increases in the National Living Wage and employer National Insurance contributions, again with "likely no material persistent effect"
  • Artificial intelligence: "AI raises productivity, and firms may reduce hiring as a result," which would mean a "structural and likely persistent increase in productivity growth"

The Bank's verdict is that "much of the recent strength in productivity growth appears to reflect cyclical adjustments." The dispersion of productivity growth across industries "is currently no higher than its pre-Covid average," and employment growth has weakened across many sectors since 2024. Those patterns point to an economy-wide cycle, not a technology concentrated in a few industries.

Chart D.3 makes the comparison directly. Productivity growth has picked up in sectors with little exposure to AI, such as wholesale, retail, accommodation and food services, as well as in information and communication, which is heavily exposed. On PPC Land's reading of the chart, information and communication output grew by around 5% a year between the turn of 2024 and the first half of 2026, while jobs in the sector shrank slightly, leaving productivity growth near 6% a year. That is fast, but the report states that it "has not risen above pre-Covid averages."

What firms told the Decision Maker Panel

The most concrete numbers come from the Decision Maker Panel, a Bank survey of company executives. "Firms responding to the DMP survey estimate that AI has increased productivity by around 0.5% over the past three years, in line with similar impacts reported in other countries," the report states. "This would point to AI contributing about 0.1-0.2 percentage points to annual productivity growth since early 2024."

Expectations are much larger. "Firms responding to the DMP survey expect AI to boost productivity growth by about 0.9 percentage points per year on average over the next three years, split roughly equally between slower employment growth and faster output growth," according to the report, which cites work by Yotzov and co-authors published in 2026. For scale, 0.9 percentage points is close to the size of the MPC's entire 1% assumption for trend productivity growth.

The Bank does not build that into its forecast. In the central projection, "productivity growth is expected to fall back over the forecast period towards its assumed trend rate of 1%." The AI gains appear instead as a risk. "But there are increasing upside risks to productivity growth from AI," the report states. "AI adoption may accelerate and could lift potential productivity growth over the forecast period and beyond, although uncertainty around the size and timing of this effect remains large." The report cites a speech by Governor Andrew Bailey from February 2026 in support.

Business intelligence points in the same direction. "Contacts of the Bank's Agents increasingly report that firms using AI proficiently and at sufficient scale can automate routine administrative work and accelerate knowledge tasks," the report states. The qualifiers - proficiently, at sufficient scale - carry weight. The Bank's evidence suggests that gains depend on depth of use, not on whether a firm has bought a licence.

The MPC says it will review its judgments for trend productivity growth and trend labour supply, "also taking into account the ONS publication of industry-level data using the new component method in February."

Jobs: fewer hires, not more firings

On employment, the report separates two things that are often blurred together. "Firms adopting AI may be able to produce more output with the same number of workers or maintain output with fewer workers," it states. Yet "in aggregate, firms responding to the DMP survey report close to zero impact of AI on employment so far."

The Bank notes that this matches academic work. It cites three studies from 2026 - by Bick and co-authors, Brynjolfsson and co-authors, and Serôdio - which "have not found clear evidence that AI has weighed on aggregate employment to date."

The distinction between hiring and firing is where the Agents' evidence adds something. "The Bank's Agents' contacts report that AI adoption is reducing the likelihood of firms hiring or replacing workers, but not materially increasing the likelihood of firing workers so far," the report states. Jobs are not disappearing through redundancies. Vacancies that would once have been filled, or departures that would once have been replaced, are being left open.

The aggregate labour market shows little sign of strain. The unemployment rate fell to 4.7% in the three months to August, according to the report, and is projected at 4.8% in the fourth quarter of 2026, rising to 5.0% by the end of 2027 and 5.1% by the end of 2028. "There have also been few signs of a pick-up in redundancies across a range of indicators," the Bank wrote. The Agents also report that "some contacts have reached their lower bound for headcount."

That aggregate calm can hide pressure on particular groups. PPC Land reported in September on Stanford Digital Economy Lab payroll data showing that employment for workers aged 22 to 25 in the most AI-exposed occupations fell 16% between October 2022 and September 2025, while older workers in the same occupations gained jobs. The same month, agency owners described how they had stopped hiring graduates as AI absorbed entry-level work. Neither finding contradicts the Bank's aggregate numbers. Both are consistent with the Agents' observation that AI is shaping who gets hired rather than who gets fired.

One detail in the labour section links directly to the advertising business. The Bank's estimate of the equilibrium ratio of vacancies to unemployment, in Chart 1.12, lists the Advertising Association/World Advertising Research Centre Expenditure Report among its sources, alongside the ONS. According to the chart notes, the model includes "the real cost of vacancy posting and hourly labour productivity" as long-run determinants of vacancies. The same AA/WARC series showed in May that UK advertising spend grew 6.4% to £46.7bn in 2025.

Memory chips and import prices

The inflation side of AI is less discussed, but the report is explicit about it. "Core goods inflation has picked up a little over recent months," it states. "Demand for AI-related products such as memory chips has placed upward pressure on export prices, particularly in China and advanced Asian economies," the report continues, citing analysis in its July edition, and this "is likely to have contributed to some of this rise."

That pressure is expected to continue. "Higher energy prices are expected to raise costs for firms abroad, while strong demand for AI-related products from China and advanced Asian economies is expected to put further upward pressure on import prices," the report states. Core goods inflation could rise to over 2.5% in the second quarter of 2027.

China is the specific worry. "Chinese export prices have risen over recent months, with increases in the prices of AI-related exports the largest factor," according to the risks section. Chinese export prices have historically been "a source of disinflation for the UK." The Bank sees a risk that they keep rising, balanced by the chance that trade diversion from changes in global tariffs pushes them down.

The projection tables show the scale of external price pressure, from all causes. UK-weighted world export prices, excluding the direct effect of oil, are projected to rise 8.0% in the year to the fourth quarter of 2026, against 4.3% expected in July. UK import prices, excluding fuel, are projected to rise 1.9%, against 0.8%. The Bank does not say how much of either revision comes from AI-related goods.

Why this matters for interest rates is set out in one sentence. "All else equal, persistently higher UK import price inflation means that the prices of domestically produced goods and services must rise more slowly for overall inflation to remain at the 2% target," the report states. A global scramble for AI hardware, in other words, leaves less room for domestic price rises, and that tightens the constraint on the MPC.

Device makers have been describing the same pressure. At its March-quarter results, Apple said June-quarter memory costs would be significantly higher, and at its June-quarter results its chief financial officer said memory explained more than all of a 120 basis point fall in adjusted gross margin. Its guidance for the December quarter, a gross margin of 45% to 46%, primarily reflects higher memory costs, according to the company.

Demand now, supply later?

The distinction between demand and supply is the core of the Bank's AI analysis. Spending on data centres, software and AI services adds to demand, and demand that outruns supply pushes up prices. AI that raises output per worker adds to supply, which allows the economy to grow faster without generating inflation.

On the evidence so far, the Bank puts the UK in the first category. "Given the observable AI-related activity in the data so far, the roll out appears to primarily reflect a boost to demand within the economy," the report states. "However, should AI investment start to feed through more strongly to UK productivity growth, that should support potential supply growth over coming years."

The timing also differs. Box C concludes that the factors behind this year's resilience - stronger global growth, AI-related activity and households saving less - will not all last. "While the boost from the AI roll-out may prove more durable, the support to UK activity from other factors is expected to wane," the report states. GDP growth is projected to slow to 1.0% in the year to the fourth quarter of 2027, as higher inflation, tighter financial conditions and fiscal consolidation weigh on activity.

So AI is, for now, a modest source of inflationary pressure on two fronts: domestic demand from the AI build-out and import prices from hardware. Its disinflationary promise, through productivity, remains a forecast held by firms rather than a fact in the data. The report does not say that AI caused the rate rise. It does list AI among the forces behind the stronger activity that the Committee cited as "adding to the case" for one.

AI inside the Bank's own models

The report also uses AI as a tool. Box A, on second-round effects from energy prices, describes a model that allows the impact of oil price shocks to vary with the state of the economy. Bank staff have extended it "to capture oil price shocks associated specifically with geopolitical events, using an AI-based measure of newspaper coverage developed by Iacoviello and Tong (2026) and Verduzco-Bustos and Zanetti (2026)."

The results suggest that slack in the labour market has so far contained inflationary pressure. But the model also warns that inflation was already elevated when the shock hit. CPI inflation was 3% in February, "close to the model's estimated 3.1%-3.5% range above which geopolitical oil price shocks have historically been associated with more persistent inflation."

What the report does not establish

Several limits apply to the AI figures.

The "just under a fifth" estimate rests on a proxy that, by the Bank's own description, includes output unrelated to AI, adjusted by a trend assumption the report does not quantify. It is a Bank staff estimate, not an ONS statistic.

The 0.5% productivity gain and the 0.9 percentage point annual expectation are self-reported by firms in the Decision Maker Panel. The report does not give the sample size, the question wording or the distribution of answers in the text. Expectations about future productivity are not measurements.

The productivity data themselves are in flux. The ONS has not yet published 2025 and 2026 figures under its new method, and the Bank relies on proxies until the November 17 release.

The report contains no breakdown for advertising, marketing or media businesses. It does not say how much of the AI proxy comes from data centres as opposed to software or consulting, and it does not quantify how much of the rise in import prices stems from AI-related goods rather than energy.

Why this matters for the marketing community

For marketers and the businesses that sell to them, the report draws a distinction that much of the industry discussion skips. The AI economy that shows up in British national accounts is the supply side of AI: software developers, data processors, consultancies and data centre operators. Advertisers, agencies and publishers mostly sit on the buying side, and that is where the Bank finds "only evidence of a modest AI-driven rise in UK productivity growth."

The finding echoes US evidence covered by PPC Land. Apollo's chief economist Torsten Slok argued in October that AI had delivered no financial payoff to US companies outside the technology sector, with margins for the rest of the S&P 500 flat near 10%. Google's own analysis in July found AI usage in 68% of occupations, with a median of only 21% of tasks done with Gemini. The Bank's 0.1 to 0.2 percentage points a year fits the same pattern: broad exposure, shallow effect.

The Agents' report that firms "using AI proficiently and at sufficient scale" are automating routine administrative work and accelerating knowledge tasks describes much of the daily work in marketing departments and agencies: reporting, campaign set-up, copy variations, research. The hiring finding is the one with the most immediate consequences for the industry's workforce. The Bank sees AI reducing the chance that a role is filled or replaced, not raising the chance of redundancy. For an industry that has relied on junior staff for production work, that is a slower and less visible change than layoffs, and harder to track in official data.

Then there is the macroeconomic setting in which advertising budgets are being planned. Bank Rate is now 4%, the market path in the report has it at 4.9% by the end of 2027, and household real incomes are projected to fall in the coming quarters. Consumption growth is projected to ease to 0.2% in the fourth quarter of 2026, and the GfK consumer confidence index fell sharply in October. Business-to-business services, by contrast, have been the strongest part of the economy, in part because of AI.

Finally, the report places the UK's AI story inside a global one that is partly financed by debt and partly priced on expectations. PPC Land has covered Ray Dalio's view that his bubble indicator puts the AI boom level with 1929 and 2000, and a Barclays strategist's estimate that AI will supply about 75% of S&P 500 earnings growth next year. The Bank's report does not address valuations. But its central claim, that the boost from AI "may prove more durable" than the other supports to growth, depends on that investment continuing.

Timeline

Summary

Who: The Bank of England's Monetary Policy Committee, chaired by Governor Andrew Bailey, with analysis by Bank staff drawing on the Decision Maker Panel survey of firms, the Bank's regional Agents and ONS data.

What: The November 2026 Monetary Policy Report estimates that novel AI-related activity accounts for just under a fifth of UK GDP growth since the start of 2026, mostly from upstream AI development and implementation. Firms estimate AI has raised productivity by around 0.5% over three years, about 0.1 to 0.2 percentage points a year, with close to zero aggregate effect on employment, though AI is reducing the likelihood of hiring or replacing workers. Firms expect a 0.9 percentage point annual productivity boost over the next three years, which the Bank treats as an upside risk. Demand for AI-related products such as memory chips is pushing up export prices in China and advanced Asian economies, adding to UK core goods and import price pressure.

When: The report was published on November 5, 2026, after an MPC meeting ending on November 4, 2026, at which the Committee voted 8-1 to raise Bank Rate by 0.25 percentage points to 4%.

Where: The United Kingdom, with reference to AI-related investment in the US and AI-related exports from China and advanced Asian economies.

Why: The Bank judges that AI so far mainly boosts demand, which adds to inflationary pressure, rather than supply, which would ease it. Whether AI's productivity gains spread beyond the companies building it will shape the UK's growth capacity and the path of interest rates. For the marketing industry, the findings show that the measurable gains sit with AI suppliers, while the effect on buyers' productivity and on entry-level hiring is still taking shape.