AI trends: agents, responsibility, inclusion
We thought it would be a useful exercise to scan the themes emerging from our member events and conversations over the past few months. Three questions come up, time and again: how do we adopt AI tools? How do we do it in a responsible way? How do we improve digital inclusion, not widen the gaps along the way?
In this month’s blog we explore the trends and collate the latest resources and discussions.
AI agents and questions of responsibility
First we had AI tools. Then we had assistants (or chatbots). Now, we have agents.
The term ‘agentic AI’, first coined in 2024, is now bandied to describe systems using AI to work more autonomously. An AI agent can take decisions and follow process flows. Think of a virtual customer service, with an AI agent answering queries, prioritising tasks, resolving problems, updating records, following up. The capabilities present huge opportunities, and also huge ethical questions.
One response to those ethical questions is emerging through funding. The National Lottery Community Fund has just announced a £3m AI for Good programme, including 50 grants to community and data ethics organisations to build a UK-wide Pulse Network, an early-warning system capturing how AI is showing up in communities: the opportunities, the harms, and the quieter shifts that might otherwise go unnoticed. It's an approach that resonates with what we're already doing through Tech4Good: the conversations in our AI Peer Network, the questions being raised by members navigating these tools.
If you’re interested in following this very live debate, look out for the All Tech Is Human workshops on the ‘agentic shift’ this July. There's a Responsible AI Mixer at SXSW London in June, and a livestream on 19 May exploring the relationship between AI, prediction and power. These conversations need voices from our community in them. Also check out the recording of a recent NESTA session on Agentic AI: opportunities and lessons for social good.
Who owns the AI your organisation depends on?
Beneath the headlines about AI capability, a structural debate is gathering pace: who controls AI infrastructure, and what does that mean for the organisations that depend on it?
When any organisation, a charity, a business, a public body uses a mainstream AI tool, how much do they really know about where their data goes or who ultimately controls it? At a time when the geopolitical relationship with the US is shifting, is it wise to depend so heavily on a handful of American tech giants? And beyond data, there's the environmental cost: a recent Guardian investigation found UK and US data centres are consuming electricity at rates that are straining national grids.
This hyperscaler vs local/sovereign debate came up in our recent AI Peer Network session. James Martin from BetterTech spoke about the real environmental cost of everyday AI use, and Robert Keus demonstrated GreenPT - a European-hosted, renewable-powered alternative that puts some of those questions in a different light. It's the kind of conversation that matters for any organisation trying to use AI responsibly, not just efficiently.
Digital skills: the inclusion gap is widening
As organisations adopt AI tools and more job roles depend on technology, the result is to further exclude those already at a digital disadvantage. It’s a tension we can’t ignore, which is why Tech4Good South West’s Digital Inclusion Network has a mighty task!
Here are some developments to inform thinking:
FutureDotNow has recently published The Ripple Effect report which focuses on the positive changes to individuals improving their digital skills.
Microsoft UK has just launched South West Connector, a new regional platform offering digital career pathways, apprenticeships, courses and an AI assistant to help people navigate options. There are also practical resources supporting people to find digital work.
The Digital Poverty Alliance’s latest research Accessibility in Digital Education Service Design highlights how digital learning platforms are failing neurodivergent learners. You can contribute to the live Government consultation Growing up in a digital world before the end of today (26 May).
The Digital Good Network has just commissioned research on 15 blueprints for what a good digital society could look like. These range from children’s and family engagement, to ‘digital death’, to combatting ageist algorithms.
Good Things Foundation AI Gateway is designed to close UK’s “AI confidence gap”.
Spark your curiosity
It’s a time of learning and opportunity, which is why we see the conversations and connections forming through Tech4Good so inspiring. Join a ‘Meet the Members’ session, find a Tech Together buddy, or come along to an event to spark build your knowledge, understand a new perspective or see a fresh opportunity.
You might also be interested in the following events or reports:
All Tech Is Human workshops on the agentic shift
Microsoft South West Connector: resources, skills development and jobs
