At the start of each year we ask funded partners to share thoughts on the big issues facing Londoners for the year ahead, covering areas like housing, migration and poverty.This year we’re focusing on one topic - AI.
Not because it's the only issue, or perhaps even the biggest one facing Londoners this year. But because AI is changing our environment rapidly, and we’re all thinking about the ways this technology may transform our society.
Chief executive's introduction
Have you seen the driverless taxis yet? In October 2025, Google's self-driving car project Waymo announced their robotaxis will be coming to London, and in the first few days of the year they’ve already been snapped driving around.
This is just one example of where AI is already reshaping our society - from the job market to how we consume information.
These new taxis aren’t just a novelty. The Financial Times has reported that London will be a battleground between Chinese and American self-driving car companies. And for London’s 121,000 black cab and private hire taxi drivers, it’s a risk to their livelihoods.
London is often where disruptive technology gets deployed first. This can be positive and exciting. But as a test-bed for innovation, London can also bear the brunt. Jobs can be put at risk or made more difficult, before we’ve got to grips with what’s going on.
We know the civil society sector is already running on empty. But if we can keep working to understand and harness AI, there's huge potential for it to improve our daily work and lives.
Crucially, we must stay alive to how it's evolving, to improve how we support low-income Londoners, and to push for policies that protect workers and communities ahead of profits.
Manny Hothi, Trust for London
Will AI lead to mass unemployment?
Throughout 2025 we saw plenty of alarmist headlines about AI, from job Armageddon's to unemployment bloodbaths. We can expect plenty more hype this year, but like most major change, the reality is more nuanced.
Evidence is mounting that AI is already reshaping the job market, but the impacts aren't being felt equally. Younger workers, 17% of who are currently unemployed, are bearing the brunt.
Our recent youth unemployment analysis explored what's driving this trend in London and what needs to change. The jobs young people typically rely on, including administration, retail and customer service, have seen vacancy drops of 40-60% in the capital over the past three years.
It's too early to put this down to AI alone, but the pattern is clear: young people and those in entry-level roles are being hit hard.
Decent work is a crucial route out of poverty. As we adopt new technology, we need to ensure that its impact on the job market, especially for those just starting off in their careers, is front and centre in policy decisions.
AI and algorithmic bias in financial services
AI is also reshaping how institutions make decisions about people. AI models are trained on existing data, and this means they can reflect existing biases.
As firms increasingly use AI for credit decisions and customer services, there are growing concerns about algorithmic bias, financial exclusion and consumer harm, particularly for those already on the margins.
People in poverty, racialised communities and other minoritised people are often forced to pay a premium (or even locked out of) the financial services and affordable credit we all rely on.
Our partners at the Finance Innovation Lab warn that the UK's light-touch regulatory approach risks repeating the mistakes of the pre-2008 era, with potentially catastrophic consequences for the people least able to weather another crisis.
Perspectives from our partners
We work with organisations all across London - experts in how things are going on the ground. Here three of our partners explore how AI is already impacting Londoners, and questions for the future.
AI and low-income workers: Amy Kinton, Organise
AI is already having a huge impact in the workplace. Amy Kinton, executive director at Organise, shares insight into some common themes the team is seeing and how they're working to support workers.
Workers, particularly in low paid or insecure work, are already feeling the impacts of AI.
In a survey of 2,000 workers in November 2025, 40% said their job had changed, got worse or they’d lost their job after AI was introduced to their workplace - but only 17% said they had been offered any input into decisions about AI at work.
Changes include extra surveillance and monitoring, for example rotas being created by AI or tracking devices that make workloads unsuitable. Other workers who are freelancers or in junior roles describe being replaced by AI tools.
Organise is helping to tackle that huge gap between AI roll-outs and worker involvement through a new project in 2026: the AI early warning system. We’ll help workers, primarily in low paid and insecure work, to evaluate AI roll-outs in real time and advocate for improvements.
The results of workers’ AI assessments will be made available, live, on a public dashboard to track AI roll-outs at major UK employers in real time for researchers and policymakers to use for national influencing too.
AI for good: Deven Ghelani, Policy in Practice
Deven Ghelani, ceo of Policy in Practice, writes about how new technology is already helping to improve lives.
AI and algorithmic decision-making tends to have a bad reputation, because it us often applied to inappropriate and harmful use cases, like fraud and collections.
But when we apply data to helping people, we find it can take the pressure off people often when they are struggling the most. At Policy in Practice we believe there is huge potential to use data much more intelligently in the benefit system.
We're leading a number of projects using data, AI and new technologies to put more money in the pockets of people on the lowest incomes:
- Our Low Income Family Tracker (LIFT) has helped streamline access to over £77m worth of support this year, by proactively identifying people missing out on over £24bn of unclaimed benefits.
- Auto-enrolment, a project supported by Trust for London, means 30,000 Londoners are now getting 50% off their water bills without lifting a finger.
- Apply Once means users can apply directly for support such as the Crisis and Resilience Fund from our benefit calculator with a single click, without having to enter their details again.
That doesn't mean we think AI is a universal force for good. But it does mean that when committed policy people apply it to solving real problems, it can help make a bigger difference.
Implications for the VCSE sector: Dan Sutch, CAST
Dan Sutch is director of the Centre for the Acceleration of Social Technology (CAST), working with funders, investors and mission driven organisations to use digital technologies to address big social challenges.
We’ve seen such dramatic adoption of Generative AI (GenAI) technologies over the past few years: famously ChatGPT grew to 100 million users in just three months. At the end of 2025 that figure was over 700 million weekly users.
Our starting point for 2026 then, regardless of whether we’re using GenAI ourselves - is to ask what are the implications if other people are using it?
As employers benefit from GenAI, we’re seeing entry level roles disappearing - current estimates are that about 1/3rd of entry level roles disappeared last year. We’re seeing huge drops in website visitors accessing specific, curated information, as search engine journey's end at GenAI generated summaries.
With the growth of AI companions, we’re seeing a shift in how people develop relationships with and through technology. And as grant seekers benefit from the use of GenAI we’re seeing overwhelm within the grant-making system.
A vital question for the VCSE sector in 2026 then is to explore the implications of these changes in behaviours, needs and expectations among the communities in which we’re embedded.
Within charity practice, we’re seeing continued adoption of AI tools - with more use within large charities with smaller organisations less supported.
The majority of adoption is looking to make existing ways of working more efficient, with some experiments exploring new ways of achieving charitable objectives by harnessing these tools - and indeed in creating tools that are owned and designed within the sector.
A clear challenge faced by those exploring AI is moving from individual experimentation to organisation alignment. Addressing this will mean managing the tension of continued delivery, evolving existing ways of working, and exploring new paradigms for creating social value.
An ambition for us for 2026 is that we can continue to develop our collective critical engagement with GenAI addressing each of these areas. To do so will require greater collaboration and shared learning.
If we can organise to do this, we can strengthen our voice, practice and influence to ensure that communities benefit from this rapid adoption.
Explore lots of helpful resources at: The CAST AI resources hub
What we can all do this year
AI has the potential to deepen existing inequalities or to help us build something better. We’re hopeful that it can be for the better.
- On a societal level, workers must have a say over how AI will impact their jobs. This is one of our core priority areas under our decent work funding programme.
- As individuals and organisations, we need to keep up with changes and be comfortable with how to use AI safely, ethically, strategically and effectively.
- And as a civil society sector, we need to be involved in conversations at a community and governmental level as technology and regulations develop. We’re proud to be part of CAST’s Charity AI Taskforce.
- Finally, government will need to work with civil society, businesses and regulators to manage the impacts of AI, rather than going all in on AI in pursuit of growth.
