What happens when one of the world's leading AI companies invites faith leaders in to help carry the weight of what it's building? And is the church ready for the conversation?
On the eve of Missional AI London 2026, James Poulter (JP) hosted the podcast's first-ever live studio recording at Premier Christian Radio in central London. He was joined by:
The conversation ranged from Anthropic's faith convenings to AI's impact on work, parenting and the global church, with questions from the studio audience.
Faith in Silicon Valley used to be something you kept to yourself. Now there's more openness, though Paul is cautious about why.
As a pastor in Palo Alto, Paul always felt the burden of helping people in his congregation who worked at major tech companies let their faith guide their decisions, and he never felt he had a great answer. In recent years, he has gone back to school and helped start an organization to help people integrate faith into their workplaces.
Silicon Valley was never exactly hostile to faith, he says; the message was just to keep it to yourself. That's changing, partly because of the political climate and some high-profile voices. Paul attended Peter Thiel's lectures and sees them as a symptom of a broader opening to faith in the Bay Area rather than its cause.
Wherever he travels, from Croatia and Paris to London, Africa and Asia, people are talking about AI. That makes it a unifying topic for the church, and Paul believes the church needs to show up in that conversation.
Anthropic has been inviting faith leaders to help it think through what it's building, and Paul came away convinced the posture was genuine.
Anthropic has hosted at least three convenings that Paul knows of: two focused on Christian groups and one interfaith gathering of around 14 religions. Without breaking any embargoes, Paul describes the tone. Anthropic opened by acknowledging that it's building something incredibly powerful, and its posture was one of openness. It felt the burden of what it's creating and wanted others to share that burden. It wasn't offering a menu of options to pick from, but there were comparisons to the Manhattan Project and a clear awareness that this could go very well or very poorly.
Is it posturing? Gareth urges a generous spirit. Social media companies did very little of this. If faith groups shape even 5% or 10% of how AI develops, that's better than nothing. Kit sees real humility in the acknowledgment that AI's impacts could be positive or negative, and she wants to encourage it. Gareth adds that people of faith are already at the cutting edge of AI with credibility in the space, unlike the church's slow start with social media.
Paul says outsiders often imagine Silicon Valley as godless and reckless. In reality, many thoughtful people, Christians and non-Christians alike, are taking these questions seriously. But the convening made clear that the forces at play are far more powerful than any one person. Companies must keep innovating and answering to shareholders, so simple fixes rarely work.
Most technologies are understood before they're built: you know what a car is before you make one. AI is different, Paul says. Builders didn't build AI; they built the thing that built AI, the trellis on which it grew. That's why a whole field, interpretability, exists to figure out what these systems are.
Gareth argues this is exactly why faith voices matter. The printing press advanced distribution and communication; social media advanced connection. AI is a development in what it means to be human, so it needs ethical and religious conversations to frame it well.
Organizations know they need AI policies. The harder work is figuring out how AI actually serves their mission.
Are churches and charities inviting technologists in? Kit sees no single response, but much more appetite than with social media, which was often left to the digital or marketing team. AI could reshape entire organizations. Many have written policies for risk and staff guidance, but then what? With little in-house capacity, charities are working together, and nearly every charity event now focuses on AI.
Kit calls her work "finding the edges": closing the gap between strategy, policy and execution. A common phrase is that an organization will use AI "where it adds value." Value to whom? That differs hugely across charities, churches and ministries. She doesn't want to turn people into AI super-users; she wants them to use technology for their charitable purpose and mission. That takes iterative, sometimes uncomfortable conversations.
Gareth warns against assuming the ministry that uses AI best will be most effective or raise the most money. AI must serve your mission. Some small churches thrive not because of their website but because of real connection. So the question is who you want to be as an organization, and how technology helps, or doesn't.
Paul sees churches wrestling on many levels. Should they put a chatbot trained on sermons on their website, an "Ask Pastor Dave" bot? How should a young youth pastor use AI, and should it write sermons? How do churches equip parents? And many people feel they have "magic wishes" in their pocket but don't know how to use them, along with fear of missing out and anxiety that they're not using AI well enough.
As AI reshapes work, the church has a role in helping people find meaning and walking with them through upheaval.
Paul jokes that someone should write a book on work and AI, a nod to JP's book AI at Work, out September 3. In Silicon Valley, workplace expectations change at whiplash speed. At Meta, he says, leaderboards once ranked employees by token usage, until it became too expensive and the company reversed course. People are left asking who they are when they're expected to manage dozens of AI agents.
We were created to do productive work, Paul says. The idea that the perfect world is one where AI agents do all our work while we relax isn't a biblical vision of work. Churches can help people recognize those lies and walk with them pastorally through the turmoil of transition.
With major layoffs at companies like Amazon, Meta, Microsoft and WPP, Paul notes that hiring and layoff cycles aren't new in Silicon Valley, even if AI now gets blamed for everything. The basics still matter: connecting with people, forming them in community and helping them through what they're facing.
Gareth points to a genuinely new pastoral challenge: what should a pastor do when a congregant falls in love with a chatbot? Paul agrees that whether it's romantic relationships or spiritual mentorship, pastors don't yet know how to respond. There will likely be shame and isolation, so churches must learn to create safe spaces where people can talk about it.
From smart speakers to self-driving cars, AI is already woven into family life, and parents are figuring it out in real time.
JP shares that he's been helping his vicar's 16-year-old son choose his A-levels, which shape university and career paths. If even a quarter of the AI predictions prove right, the world will look radically different in seven years, and we're ill-prepared for those conversations.
Gareth's seven-year-old is already learning about AI at school. When Alexa gave her a wrong answer, she was shocked. "But Alexa said it." It was an important lesson that the voice in the kitchen isn't always right. He jokes that his daughter never thought he knew everything, but she might think Alexa does.
Kit, a mother of young twins, finds AI helpful for navigating the flood of school apps, information and family decisions, but stays alert to what's being surfaced. Her curious daughter already says, "Ask Google," and Kit can see that becoming "Ask AI." So she and her husband are thinking carefully about the role of technology in their family and the critical thinking their kids will need.
Paul's experience is different. His children, in their early twenties, are less interested in AI than he is and tease him for using it. They're digital natives but not AI natives; it arrived while they were at university, where they were told not to use it. When one had car trouble and Paul suggested asking Claude, the reply was that it might be wrong, so they'd have to research it anyway. Meanwhile, Paul rarely touches his steering wheel, thanks to self-driving cars. Gareth recalls a video call where Paul was being driven by a driverless car, which unnerved him the entire call.
Kevin from Premier asked: if people who grew up with sat nav never learned to read a map, will AI leave us dependent and missing core skills?
Paul acknowledges the risk, but notes he also gained something from digital maps: a constant sense of where he is in the world. We'll surely lose something with AI, but we may not yet know what we'll gain. Gareth urges organizations not to delegate thought leadership to AI; they still need to think and have convictions, or they become a shell. He's also grateful technology lets his agency serve the US, UK and Australia, something impossible 20 years ago. Kit adds that the purpose isn't to read a map; it's to reach your destination. If something makes you feel alive, like creating art, keep doing it. If AI cuts admin so you can serve your mission better, adopt it.
Asher, a 20-year-old founder, asked how the next generation can make sure AI serves everyone, not just a few powerful founders. Kit notes that many large models are trained mostly on data from the Global North and miss local cultures, idioms and metaphors. In the humanitarian AI-for-good space, there's growing interest in local and sovereign AI models that are contextual, could shift power dynamics and may be more environmentally friendly.
Paul frames AI as a tool we must learn to use well. Put a hammer in a room of middle schoolers and they'll probably hit each other with it. "AI is a great librarian but a bad teacher," he says. In Bible study, he asks what scholars say about a word or for a survey of patristic thought, not simply what a verse means.
JP adds that the bigger gap is public education. Governments and industries are waiting for AI companies to provide answers, but there's no commercial incentive for them to do so. Safety has always depended on public education, and part of Missional AI's mission is to provide that for the church and faith community.
Chante, a data manager with Biblica in South Africa, asked how Africa can close the gap in AI adoption and education. Many people, she noted, think AI simply means ChatGPT.
Gareth points to Africa's projected economic growth and to the shifting center of global Christianity. He cites an often-repeated claim, which he attributes to Lausanne and admits needs fact-checking, that the average evangelical is now a 14-year-old Nigerian girl. That changes the shape of the church and of technology.
Paul notes that technology follows economics. Social media's problems weren't really technological; they were driven by its business model. Where markets grow, technology follows. JP adds that Western concerns, like AI in schools or mental health care, may look very different where an AI educator could offer far more than people currently have access to. Kit sees a parallel with renewable energy: places without existing infrastructure can leapfrog straight to new solutions, as Kenya did with mobile banking through M-Pesa.
Watch the full live episode on the Missional AI Podcast, and let us know in the comments if you'd like more live recordings.
At missional.ai, you can sign up for the newsletter, join the Slack community, explore talks from Silicon Valley, London and past gatherings, and register for upcoming events, including the next Global Summit in San Francisco, March 30–April 1, 2027.
Thanks to Premier Christian Radio for hosting this live recording.