SINGAPORE: Over the past year, artificial intelligence (AI) has been everywhere – in policy speeches, boardrooms, university seminars and dinner conversations.

At some point, repetition creates fatigue. Not because AI is unimportant, but because the conversation can start to feel generic. When everything is about AI, nothing feels concrete.

That is why the announcements in Budget 2026 are more significant than they appear to be.

The decision to establish a National AI Council chaired by Prime Minister Lawrence Wong, together with selected “AI missions” in advanced manufacturing, connectivity, finance and healthcare, marks a shift from aspiration to execution.

This is not simply about declaring AI a national priority. It is about choosing where Singapore intends to compete and organising the system to deliver.
 

DEFINING A DIRECTION

The selection of the four sectors is deliberate.

Advanced manufacturing reflects our strengths in precision engineering and high-value production. Connectivity reinforces Singapore’s position as an air and sea hub, where reliability matters as much as innovation. Finance builds on a globally-connected financial centre where regulatory credibility is itself a competitive asset. Healthcare responds to demographic reality, while leveraging our strengths in clinical excellence and systems management.

By naming sectors and framing missions, the government is signalling that AI will not be pursued as a diffuse overlay across everything. It will be deployed deliberately where productivity gains and exportable capabilities can be meaningful.

The more interesting question is what the new AI Council must actually do.

Chaired at the highest level, the council recognises that AI is not a narrow technology issue. It cuts across issues ranging from manpower to education, regulation, data governance, cybersecurity, procurement and public trust.

Without coordination, agencies optimise locally and pilots fail to scale. The value of the council will not lie in producing another strategy document. Instead, it will lie in removing bottlenecks and aligning incentives.

INTEGRATION IS KEY

When it comes to AI, the hard part is not experimentation. It is integration.

That means addressing less glamorous constraints – fragmented data systems, limited interoperability, unclear procurement pathways and uncertainty around governance. Common architecture is what will turn AI from a collection of promising pilots into systemic innovation.

Prime Minister Lawrence Wong acknowledged that end-to-end AI transformation demands shared foundations, not piecemeal initiatives. That logic is central to proposals such as the Networks for Humanity Research Institute at the National University of Singapore (NUS), which seeks to establish open standards so financial institutions can modernise collectively, rather than in costly and incompatible silos.

In finance, in particular, the integration challenge is becoming structural. As AI evolves from analytics tools into autonomous agents allocating capital, executing compliance and interacting with markets, the underlying infrastructure must adapt. In this case, programmable, quantum-secure and interoperable systems will be prerequisites for stability in AI-driven markets.

TURNING AI ACCESS INTO PRODUCTIVITY GROWTH

The Budget also includes measures to spur adoption, including six months of free access to premium AI tools for Singaporeans who take up selected courses, alongside productivity support for firms.

These are sensible entry points that reduce friction and encourage experimentation, but access to tools does not automatically translate into productivity. The real gains come when organisations rethink workflows, redesign processes and integrate AI into core operations.

Many firms discover that the binding constraint is not access to a model. It is data quality, system integration and organisational readiness. Without addressing these, AI will remain an interesting add-on, rather than a transformative capability.

At the individual level, the same principle applies. Prompting skills are useful but durable advantage comes from combining domain expertise with AI fluency. The worker who understands both the workflow and the tool will outperform the one who simply knows how to operate the interface.

In line with this logic, discussions are underway for a new master’s programme at the NUS Business School. With a focus on AI in business, this will give professionals a structured way to build strategic capabilities to work with and alongside the technology.

This is why the deeper strategic question is whether Singapore can convert early AI adoption into sustained productivity growth – a necessity for a mature economy facing demographic constraints.

If deployed effectively in manufacturing, AI can raise output without proportionate increases in labour. In finance, it can strengthen risk management and compliance, while improving customer experience.

AI can help manage rising demand in healthcare without overwhelming human capacity, while in connectivity, it can reinforce our role as a trusted node in global networks.

With a prime minister-chaired council showing priority and sector missions signalling focus, the next thing to watch will be measurable outcomes. Are turnaround times falling? Are error rates declining? Are new AI-enabled products being exported? Are workers being redeployed into higher-value roles rather than displaced into uncertainty?

If Singapore can combine coordination, institutional capacity and speed, this may well be remembered not as another AI headline, but the moment when ambition became advantage.

Sumit Agarwal is the Low Tuck Kwong Distinguished Professor of Finance, Economics and Real Estate at the National University of Singapore (NUS) Business School, and the managing director of the Sustainable and Green Finance Institute at NUS. The opinions expressed are those of the writer and do not represent the views and opinions of NUS.

SINGAPORE: A new National Artificial Intelligence (AI) Council will be established and chaired by Prime Minister Lawrence Wong to coordinate and drive Singapore’s AI strategy.

The council will oversee the development and execution of “AI missions”, said Mr Wong on Thursday (Feb 12) in the Budget 2026 statement.

“These missions will drive AI-led transformation in key sectors of our economy, and push the boundaries of what is possible for Singapore and for the world,” said Mr Wong, who is also the finance minister.

The missions will focus on four sectors: advanced manufacturing, connectivity, finance and healthcare.

“For example, in advanced manufacturing, we aim to accelerate innovation and build best-in-class factories that can compete globally,” he said.

“In connectivity and logistics, AI can help to automate airport and seaport operations, move goods more efficiently and strengthen Singapore’s position as a leading global hub.”

He said the AI missions are not “abstract aspirations” but come with clear objectives and tangible outcomes.

But delivering these objectives will “require us to work differently”, he said.

“Within the government, we will better align our R&D (research and development), regulatory and investment promotion efforts, so that agencies act in concert and pull in the same direction.”

This will require a coordinated national effort, he said, and the establishment of the National AI Council will “provide strategic direction and to drive Singapore's AI agenda”.

The council will comprise:

• Deputy Prime Minister and Minister for Trade and Industry Gan Kim Yong
• Coordinating Minister for Social Policies and Minister for Health Ong Ye Kung
• Minister for Digital Development and Information Josephine Teo
• Minister for Manpower and Minister-in-charge of Energy and Science and Technology Tan See Leng
• Minister for National Development and Deputy Chairman of the Monetary Authority of Singapore Chee Hong Tat
• Acting Minister for Transport and Senior Minister of State for Finance Jeffrey Siow

FEAR OF AI CANNOT BE THE RESPONSE

While the potential of AI to raise productivity, unlock new discoveries and transform lives is immense, this promise also comes with deep concerns, such as job displacement, misinformation and the ethical use of powerful technologies.

These anxieties are real and must be confronted squarely, said Mr Wong.

“But fear cannot be Singapore’s response,” he said. “If we allow uncertainty to paralyse us, we will fall behind in a world that is moving rapidly ahead.”

He said how AI is developed and used in Singapore will be defined, with clear rules set to ensure it is applied responsibly and safely.

“And we will ensure that its benefits are shared widely across society,” he said.

He added that AI can help Singapore overcome structural constraints such as limited natural resources, a rapidly ageing population and a tight labour market.

However, Singapore’s advantage does not lie in building the largest frontier models, but in deploying AI effectively, responsibly and at speed.

“Singapore can be a trusted hub where companies and researchers come together to develop, test and deploy impactful AI solutions, and do so faster and more coherently than many larger countries,” he said.

He noted that companies such as Google and Microsoft have set up AI centres of excellence in Singapore, creating a growing number of jobs for Singaporeans.

But to fully realise AI’s potential, Singapore needs to move beyond individual pilots and isolated experiments.

“We must organise at a national level, and move with speed and scale,” he said.

NEW AI PARK

To bring more companies on board with AI transformation, a new Champions of AI programme will be launched to support firms with the ambition to use AI to comprehensively transform their businesses.

Support will be tailored to each company and will include enterprise transformation and workforce training.

“As these companies succeed, they will set benchmarks for their industries and inspire others to follow,” said Mr Wong.

Support for all enterprises, especially small- and medium-sized enterprises, will be strengthened.

The Enterprise Innovation Scheme will be expanded to include AI expenditures as a qualifying activity for the assessment years of 2027 and 2028, capped at S$50,000 (US$39,600) per year.

The scheme currently provides businesses with 400 per cent tax deductions on qualifying expenditures in activities such as research and development, innovation and capability development.

The Productivity Solutions Grant, which helps companies adopt digital solutions, will also be expanded to cover a wider range of digital and AI-enabled solutions.

A new AI park will be established at one-north. It will build on a pilot initiative called Lorong AI, a dedicated co-working space for the AI community at Cross Street.

“This will be a new cluster to catalyse ideas, forge collaborations, and translate AI initiatives into practical solutions for businesses and public services,” said Mr Wong.

"SINGAPORE IS WELL POSITIONED TO LEAD"

Major AI firms say that Singapore's AI plans are timely. 

Mr Oliver Jay, managing director, international at Open AI, said that the economic gain from AI lies in closing the "capability overhang", which is the gap between what AI can do and how it typically is used.

"Closing that gap requires better tooling, training, and governance so businesses can safely deploy higher-value use cases," said Mr Jay. 

He added that Singapore is in a good position to lead, ranking high for per-capita ChatGPT adoption, which is Open AI's AI chatbot. 

"We look forward to partnering closely with the government, the upcoming National AI Council, and industry players to translate capability into measurable economic impact," he said.

Ms Eileen Chua, managing director of business software and AI firm SAP Singapore, said her company’s research shows that Singapore firms invested an average of S$18.9 million in AI over the past year, generating an average return of 16 per cent.

That figure is expected to rise to 29 per cent within the next two years.

"This illustrates that AI is already moving beyond concept into business impact," she said. "Government support that enables clear objectives, regulated sandboxes and coordinated industry missions will help enterprises accelerate that transition from early returns to sustained value." 

SINGAPORE: Chocolate manufacturer Barry Callebaut has opened its first global innovation centre outside of Europe in Singapore, tapping artificial intelligence to drive growth in Asia.

The Zurich-headquartered company said its new facility at the Singapore Science Park will also boost research and development efforts and support the creation of products tailored for fast-growing Asian markets.

It brings together the world’s first AI centre dedicated to chocolate and cocoa, alongside a cacao coatings centre focused on product development, according to a media release on Tuesday (Feb 10).

The site also houses the Callebaut Chocolate Academy Singapore and a regional R&D facility, allowing the company to translate consumer trends and customer briefs into new concepts before testing them at a new pilot lab at its Senoko factory.

“Singapore is a hub for innovation and (an) artificial intelligence ecosystem,” said Mr Vamsi Mohan Thati, the company’s president for Asia Pacific, the Middle East and Africa.

“What we hope to do is to bring customers to this innovation centre, co-create product concepts and new ideas with them, and then take them to commercial scale using the pilot lab,” he added.

INNOVATING FOR TROPICAL CLIMATES

Barry Callebaut is a major player in the chocolate and cocoa industry, producing ingredients used in a range of products such as ice creams, fillings and coatings.

The company has had a presence in Singapore since 1997, when it established its Senoko plant – now one of the region’s largest industrial chocolate manufacturing facilities.

What began with 60 employees has grown to more than 300 staff across commercial operations, manufacturing and R&D.

The new centre will host more than 30 specialised roles, including engineers, food scientists and chefs, with plans to double the team size over the next three to five years.

Among its priorities is developing products suited for Asia’s climate and consumer preferences – including chocolate that melts less easily in tropical heat.

Mr Thati said the team in Singapore has also developed a chocolate concept with higher levels of flavanol – naturally occurring compounds in cocoa linked to antioxidant properties – and that the product is currently being test-launched in China.

The centre allows the company to support large and rapidly expanding markets such as China, India and Indonesia and the wider Southeast Asian region, Mr Thati said.

Beyond new chocolate formulations, it will focus on helping customers manage pricing pressures.

“They both want products with innovation, but they also want us to innovate on current products to keep the cost under control,” Mr Thati noted.

“Pricing is important in this region, and price points need to be maintained by our customers, so we work with them to make sure that we are innovating, not just on product, but also on the cost side.”

BUILDING TALENT PIPELINE

The company says it plans to expand the facility into a larger, more connected ecosystem – one aimed at creating more than just chocolate.

Speaking at the facility’s opening ceremony on Tuesday, Minister of State for Trade and Industry Gan Siow Huang said Singapore has built a strong food innovation and AI ecosystem to support companies amid intensifying competition and shifting consumer tastes.

The country plays a key role in the global cocoa market, handling around 15 per cent of worldwide trade flows, she noted.

It is also home to more than 200 agri-food tech startups working on areas including alternative foods, functional ingredients and precision agriculture.

As technological disruption reshapes industries, Singapore continues to strengthen its talent pipeline in emerging and in-demand areas such as R&D and AI, Ms Gan added.

Singapore aims to triple its pool of AI practitioners to 15,000 in the coming years.

Company-led efforts, such as Barry Callebaut’s innovation centre, complement this push to deepen skills in such emerging areas, she said.

SHANGHAI/BEIJING: At an art exhibition held in Shanghai, visitors were able to get a taste of what creativity looks like in an age where artificial intelligence is both a tool and a collaborator.

Many of the artists featured in the exhibition, titled PromptoScape, were not human but machines.

While the event – which took place from July to September last year – has ended its run in the Chinese megacity, court cases over AI-assisted creations are only just beginning.

Legal battles over such creations are brewing in courtrooms across China, as judges and lawyers race to keep pace with rapidly evolving technology.

Chinese AI models have been making global headlines for rivalling the best in the world, with many Chinese creative professionals adopting such technology amid government promotion of AI integration.

A study by Shanghai Jiao Tong University, published two years ago, found that 87 per cent of about 110 visual artists who were surveyed use AI in at least one stage of creation.

This has planted the seeds for a new legal conundrum: Does AI artwork generated by artists warrant copyright protection, especially when they are created using models trained on vast datasets of existing creative works?

A LEGAL WIN AT A COST

One artist caught at the centre of this legal grey zone is Lin Chen, known professionally as Tudou Man.

In 2024, he found himself at the forefront of a landmark lawsuit involving AI-assisted art.

He had used a generative AI tool called Midjourney to “do an early round of generation” of an image.

“After a few hours of back-and-forth, a heart finally appeared lying on the surface of Shanghai’s Huangpu River. I then edited it in Photoshop, turning it into half a heart on water,” he told CNA.

“This brings out the idea: when the two halves meet, it’s a complete heart.”

But Lin later saw a video online showing a 3D installation of his artwork, displayed on a lake in neighbouring Jiangsu province.

“It was extremely infuriating,” Lin recalled.

At first, Lin tried to reach out to the real estate company, but they did not respond.

Then he decided to contact the balloon installation company – which, besides creating the 3D balloon installation, also used an image similar to his artwork for their online advertisement.

The company responded but was adamant they did nothing wrong.

“They said they had created many half-suns and half-moons on the water before, so why would half a heart be an infringement? They didn’t believe using my work constituted an infringement,” Lin said.

“That’s why I filed the lawsuit.”

Lin ultimately won his case, but with an unexpected twist.

In a 27-page ruling – an unusually lengthy one for a civil case – the Changshu Court in Jiangsu decided that the 2D image used in the balloon company’s advertisement infringed Lin’s copyright.

The court found that even though Lin created the image using an AI tool, he demonstrated originality due to his creative control in the composition of elements like colour, environment and light.

However, the court ruled that the 3D installation on the lake was not a copyright infringement.

In its ruling, the court stated that copyright protects concrete artistic expression rather than abstract creative ideas, and the heart-shaped balloon installation represented a common and simple design concept that lacked sufficient originality.

The also court pointed out that Lin’s copyright registration only covered the 2D artwork itself.

Lin was awarded 10,000 yuan (US$1,400) in damages for the illegal reproduction of his 2D artwork. But he had spent more than seven times that amount on legal fees alone.

He said it was a learning experience for him in how to register his artwork for copyright protection. He registered it as a single, flat two-dimensional image, but noted he should have applied for an appearance patent to copyright it from multiple angles.

“If it had been registered as a 3D piece or in other sculptural forms, it wouldn’t have left loopholes for the other party – or anyone else – to exploit,” he lamented.

CHANGE IN LEGAL THRESHOLD

Meanwhile, in the Chinese capital Beijing, lawyer Deng Yile is defending the copyrights of another AI-assisted artwork.

Her client, who created the piece with the help of AI, has accused a Chinese cybersecurity firm of using a marketing image that bears a striking resemblance to his creation.

The case has, however, taken a turn – Deng’s client found himself being sued for defamation instead.

Deng and her legal team had to debate what course of action to take, after Chinese courts ruled last year that a creator who enters AI prompts or adjusts parameters will not secure copyright protection for their AI-generated image.

This was a reversal of the courts’ decision in 2023 and reflects how judicial rules are responding to tech development, said Deng.

Her legal team began exploring a new strategy - to frame AI-assisted artworks as unique datasets protected under recently revised civil provisions issued by China’s Supreme People’s Court.

Those provisions set out several new courses of action to protect the intellectual property of data.

Deng’s team is arguing that by using data without permission to generate their own artwork, the defendants have infringed on her client’s rights.

But this has raised a larger, unresolved issue. 

AI models themselves are trained on enormous datasets of creative works, often without compensation, under the principle of fair use.

Still, Deng said she feels the fair use principle is worth re-examining in the current AI age.

“If we cannot draw a clear line for what counts as ‘fair use’, that lack of clarity will also affect (AI) development,” she pointed out.

“We can see there are a large number of works in the public domain. But if AI companies ‘learn’ from them, or even conduct style-based learning, has that already gone beyond what is necessary for (fair) use?”

The principle of fair use was also mainly to regulate personal use for learning purposes, noted Deng, but AI platforms or large learning models use data for primarily commercial purposes.

NO EASY ANSWERS

Globally, the debate is intensifying.

In 2024, more than 10,000 actors, musicians and authors around the world signed a public statement warning that unlicensed use of their work to train generative AI poses a “major, unjust threat” to their livelihoods.

The number of signatories later grew to 50,000, including novelist Kazuo Ishiguro, actress Julianne Moore and singer-songwriter Kate Bush.

While their demand to stop unlicensed AI training altogether appears unlikely to gain traction, amid the international race for AI dominance, others have suggested people should treat creative knowledge as a public good and collectively fund it like public broadcasting or infrastructure.

For Lin the artist, his experience has reshaped how he works.

Rather than rejecting AI, he had embraced it early, believing the technology had to be mastered, not feared.

Now a practising designer and university lecturer, he has watched entire courses disappear as AI replaced once-essential technical skills.

Today, he is known for AI-generated digital art that remixes modern consumer culture with Chinese aesthetics – hyper-realistic, imaginative works that sometimes resemble fan art of major brands.

Some of those brands have even approached him directly, paying to use his artwork for their marketing materials or to generate more conceptual artwork for them.

After his lawsuit, fellow creatives have begun seeking his advice on how to protect their own work.

As China pushes ahead with AI as a national strategy and millions of AI-generated creations flood the market, the rulings emerging from its courts may offer an early glimpse into how ownership, originality and creativity could be redefined – not just in China, but around the world.

SINGAPORE: Despite government projections that economic growth would slow as Singapore's economy matures, the country has defied expectations for two consecutive years.

In 2025, Singapore's economy expanded 5 per cent, exceeding forecasts even after the Ministry of Trade and Industry (MTI) upgraded its projections twice – in August and November. The year before, growth hit 5.3 per cent.

This marks the first time since 2010 and 2011 that Singapore has sustained annual growth above 5 per cent for two straight years.

On Tuesday (Feb 10), MTI raised its 2026 forecast to between 2 and 4 per cent, up from the previous forecast of 1 to 3 per cent.

CNA examines what drove this exceptional performance and whether Singapore can maintain its momentum.

How did Singapore do so well?

MTI attributed 2025's robust GDP expansion primarily to manufacturing, wholesale trade, and finance and insurance sectors.

The electronics cluster and machinery, equipment and supplies segment experienced particularly strong growth, driven by surging demand for artificial intelligence-related electronics. 

Accommodative financial conditions supported growth in the finance and insurance sector.

Ms Selena Ling, chief economist and head of group research at OCBC, highlighted "very broad-based sectoral growth", citing robust foreign direct investment and safe haven capital inflows into Singapore.

Beyond semiconductors, construction remained resilient with strong public and private sector pipelines, while pharmaceuticals also contributed significantly, she added.

Singapore also benefited from "competitively lower" tariffs on exports to the United States compared to other Southeast Asian economies, Ms Ling noted.

Standard Chartered economists said relatively benign global monetary and fiscal conditions, robust AI-related demand, tariff truces and lower effective tariff rates likely combined to boost economic activity.

How unusual is this performance?

Economists agreed that Singapore's GDP performance over the past two years has been exceptional for a developed economy.

In recent history, Singapore has only recorded such high growth rates when recovering from major crises, said Ms Lee Yen Nee, senior country risk analyst at BMI Research.

"Last year, this growth rate was achieved without a preceding crisis, and it was almost entirely driven by the global capital spending on AI," she said, adding that while electronics manufacturing benefitted most, spillover effects reached services sectors involved in goods movement.

Ms Sheana Yue, senior economist at Oxford Economics, called it "particularly exceptional" that Singapore's economy expanded by 5 per cent despite "heightened external challenges amid US tariff hikes".

In April last year, MTI downgraded its GDP forecast to between 0 and 2 per cent on fears that US tariffs would severely impact Singapore.

Ms Lee said Singapore appeared largely shielded because semiconductors, pharmaceuticals and many high-tech products remained exempt from tariffs.

"Singapore’s high exposure to these areas means that the economy is largely shielded. Furthermore, Singapore benefitted from the flurry of front-loading activity to beat the implementation of higher US tariffs," she added.

Did other Asian economies see similar growth?

Several Asian economies also recorded strong growth in 2025, economists noted.

Taiwan's economy grew 8.63 per cent – its fastest pace in 15 years – while Malaysia's advance estimates indicated 4.9 per cent growth.

"The main driver was exports and investment related to electronics, AI in particular," said Ms Yue, describing Taiwan and Malaysia's performance as "stunning".

The surge in demand for AI products, coupled with tariff-related front-loading, benefited the wider Asian region where high-end chips are produced, she said.

Ms Yun Liu, senior ASEAN economist at HSBC, pointed out that Singapore's growth was almost the same as what Indonesia reported, and that Vietnam grew around 8 per cent.

Why did forecasts underestimate growth?

Singapore initially underestimated AI-related electronics demand, acknowledged Ms Yong Yik Wei, chief economist at MTI, during a media briefing on Tuesday.

"We underprojected because it's quite a nascent technology, and it's quite hard to get the exact trajectory with a high level of precision. So on hindsight, we did underproject that strength in the AI demand," she said in response to CNA's question about the ministry's two forecast upgrades last year.

That demand created positive spillover effects into related sectors such as wholesale trade, she added.

Tuesday's forecast upgrade takes into account "sustained momentum" in AI investment, aligning with the Economic Strategy Review committees' recommendations for Singapore to establish itself as a global AI leader.

MTI also underestimated pharmaceutical output due to unexpectedly large production of a high-value active pharmaceutical ingredient, Ms Yong said.

"It's actually quite difficult to forecast pharma output because it's quite volatile because of the nature of the production process," she said.

Can Singapore sustain this growth?

Analysts caution that maintaining similar growth in 2026 will be challenging, particularly given 2025's high baseline.

"If Singapore’s AI hub ambitions can materialise, then it is possible to see slightly better-than-expected growth potentially exceeding 3 per cent ... but 5 per cent may be a bit of a stretch," said Ms Ling of OCBC.

Ms Yue of Oxford Economics pointed to emerging slowdown signs, noting fourth-quarter GDP growth declined from the third quarter.

External sector risks, uneven domestic activity and a softening labour market could weigh on private consumption momentum, she added.

However, the Singapore Economic Development Board (EDB) remains optimistic about AI's prospects.

"One of the strengths for Singapore is really in our industry verticals. We have various strengths in key sectors, for example, semiconductors, in electronics, in healthcare and so on, that provides for that fertile ground for AI (application)," said Ms Jillian Lim, executive vice-president at EDB.

With leading tech companies establishing operations in Singapore, she expects AI to remain an important growth area.

"We do expect that AI will be a key pillar and a growth area for us that will sustain," she said.

LONDON: Visions of an AI-infused world can be a little scary. Perhaps our brains will dull as we outsource intellectual struggle to our digital assistants. Perhaps – brace yourself – your jaunty economic analysis will come from a confident large language model, rather than a harried human.

I prefer to daydream about a sunnier scenario, in which our new digital tools deliver huge productivity gains. So perusing the latest data and evidence, where are the glimmers of light?

One lies in the excitingly strong headline labour productivity growth in both the UK and US. Admittedly, my threshold for excitement here is quite low, though more importantly there are easier explanations for the uptick than an AI-fuelled boom.

In the US, tariff uncertainty could have made companies hesitant about hiring, while in the UK a higher minimum wage may have been clearing out low-paid jobs. Both could raise measured productivity, but not really in the way we want.

WHAT THE DATA SAYS

More promising signs came from digging into the more granular data. I don’t mean the assortment of corporate anecdotes that have the informational value of parental boasting at the playground.

One compilation by Goldman Sachs put the average productivity boost from AI at 32 per cent. And based on the playground chatter, the average child is a chess prodigy, a lover of Swiss chard and a semi-professional trombonist.

Instead, I mean correlations at the industry level. If AI were helping companies squeeze more out of their employees, you would expect the industries adopting AI most enthusiastically to be enjoying the strongest labour productivity growth.

In the US that correlation has started to show up in recent data. Though of course, correlation isn’t causation and it could be that more innovative industries were most likely to adopt AI in the first place.

In a post published by the Federal Reserve Bank of St Louis, some economists try to improve on this analysis in two ways. First, rather than blunt AI adoption metrics, they ask people to estimate how much time AI tools saved them at work.

Second, they look at recent labour productivity growth between the introduction of ChatGPT and the second quarter of 2025, relative to its trend between 2015 and 2019. This was supposed to strip out any pre-existing trends that could mess up the results.

Combining these two metrics, they found that the industries where workers were saving the most time using AI were also the ones seeing unusually fast labour productivity growth. These included information services as well as professional, scientific and technical services. And updating the data to the third quarter of 2025, it looks like the correlation strengthened slightly.

I wouldn’t take the self-reported time savings too literally, not least because not everyone is as diligent as me, reallocating the time I save using ChatGPT (to find data) towards making my output even more jolly. If some people use the extra time to perfect passive-aggressive emails to their colleagues, again, that’s not really the kind of change we want to see.

A FINAL BRIGHT SPOT

It’s also reasonable to be sceptical of these correlations because LLMs have only recently graduated from “precocious 11-year-old” to “cocky graduate intern”, and towards the end of 2025 self-reported AI adoption by US businesses was still below 20 per cent. So my final bright spot comes in the form of a study taking a longer view of the data and a broader view of the technology.

Jonathan Haskel, one of the study’s authors, explained that 2017 was the real technological turning point, when a famous “deep learning” paper introduced the transformer architecture in machine learning (the “T” in ChatGPT), boosting generative AI. Which is why they compare the period between 2017 and 2024 with the one between 2012 and 2017.

More specifically, the authors study US investment in software and estimate how much it has contributed to growth. This involves various assumptions, as they try to include both the productivity gains associated with companies becoming better at producing the software, as well as the effects of other industries using it. They estimate that together, these contributed as much as half of the increase in productivity growth between the two time periods.

All of this is suggestive – the sun clearly hasn’t come out fully yet. We don’t have the data to repeat that last analysis in Europe. And when senior McKinsey adviser Tera Allas examined the British data, she couldn’t find any evidence that AI-adopting industries were experiencing unusually high productivity growth.

Still, I’m trying to stay positive. Otherwise maybe you will decide that your columnists are better in artificial form.