AI in geospatial: disruption and opportunity - 2025 webinar review
The Royal Geographical Society's Small, Micro and Medium Enterprises Professional Practice Group hosted a webinar titled AI in Geospatial: Disruption, Opportunity and What Comes Next, chaired by Andy Murdock, Director at Maploom and Acting Chair of the RGS SMME Group.
The session brought together experts from across the geospatial and environmental technology sectors to explore the accelerating role of artificial intelligence (AI) within geography, Earth observation, and data-driven decision-making.
Speakers included Dr Iris Kramer, Founder of ArchAI; Alexandra Simmons, Founder of Oko AG; and Dr. Alastair Graham, Director of AJG Consulting.
Together, they examined how AI is being applied in practice, its opportunities and challenges for the geospatial profession, and the wider implications for data ethics, skills, and environmental sustainability.
You can watch again and view the slides from the webinar, follow the links at the bottom of the page.
Setting the Scene: AI’s Growing Influence in Geospatial
In his introduction, Andy Murdock outlined the evolution of AI within the geospatial discipline. Geographers are very much early adopters of AI and are well placed to exploit AI, having long used machine learning to classify imagery and detect change.
However, recent years have seen a major leap forward with the emergence of GeoAI—the integration of AI directly into GIS platforms—and now generative AI, powered by large language models such as ChatGPT.
He noted that the increasing sophistication of AI tools has already begun transforming how spatial data are processed and interpreted. Examples include AI-assisted classification, feature extraction, automated change detection, and even conversational interfaces within software such as Esri’s AI Assistant and the Spatial Analysis Agent in QGIS.
Andy also observed that while much of this technology remains in preview or beta form, the volume and richness of geospatial data, combined with the repetitive nature of many spatial tasks, make the sector well suited to AI-driven innovation. At Maploom, for instance, AI is already being used to summarise documents, assist with coding, and configure client platforms more efficiently.
However, he cautioned that the industry faces questions over cost structures (credits / tokens), licensing models, and environmental impacts associated with large-scale AI deployment. These themes set the stage for the expert contributions that followed.
Dr Iris Kramer — Founder, ArchAI
Iris presented the work of ArchAI, a company born from her PhD research on using AI to detect archaeological features from Earth observation data. ArchAI applies image recognition and machine learning to identify features in satellite imagery, LiDAR data, and historic maps—ranging from burial mounds and ancient field boundaries to lost ponds and woodland habitats.
The technology supports land development, risk assessment, and nature recovery by providing consistent, national-scale insights into the historic environment. Clients include the Forestry Commission, utilities, and local authorities.
Iris emphasised that while AI can automate the discovery of spatial patterns, expert interpretation and rigorous quality control remain vital. Overconfidence in model outputs or poorly trained crowdsourced datasets can produce misleading results.
She cautioned against the “black box” tendency of AI and underscored the importance of expert validation and transparency in model design and data provenance.
Alexandra Simmons — Founder, Oko AG
Alexandra Simmons introduced Oko AG, an early-stage technology company using AI and geospatial data to simplify access to agricultural funding. She highlighted that more than 1,000 public and private funding schemes exist across the UK, often described as “impenetrable,” with complex eligibility criteria and policy language.
Alexandra’s platform combines a national funding directory with geospatial intelligence and conversational AI. Farmers enter a unique identifier, which is cross-referenced against geospatial data and tagged funding schemes to identify opportunities specific to their land and soil type. A chatbot interface allows users to query the data conversationally—eliminating the need to interpret long policy documents.
Alexandra explained that Oko uses AI agents with task-specific prompts, designed in collaboration with agricultural experts, to ensure accuracy and accountability. She also described Oko’s strong commitment to user accessibility, incorporating neurodivergent design principles given that around 36% of farmers identify as dyslexic or having ADHD.
She acknowledged the challenges of user trust and early adoption, noting that human oversight, quality control, and transparent feedback loops are essential to ensure reliability in AI-assisted decision tools.
Dr Alastair Graham — Director, AJG Consulting
Alastair Graham, an experienced Earth observation consultant and co-host of the GeoMob podcast, reflected on the broader ethical and societal implications of AI. He drew parallels with previous technological revolutions, warning against “uncritical adoption” driven by major technology corporations and the need to ensure equity in the distribution of the benefits from AI.
Quoting from Blood in the Machine by Brian Merchant, Alastair reminded the audience that historical “Luddites” were not anti-technology but resisted tools used against workers’ interests. He urged the geospatial community to ensure that AI development remains transparent, equitable, and open, aligning with the ethos of open-source geospatial software promoted by OSGeo UK.
Alastair also highlighted the environmental and social dimensions of AI’s expansion. The construction of global data centres—forecast to attract £2.2 trillion in investment by 2029—raises sustainability and resource questions. He called for geospatial professionals to advocate for responsible AI that supports both planetary and societal wellbeing.
Panel discussion: challenges, ethics, and skills
In the panel discussion that followed, several cross-cutting themes emerged:
- Trust and transparency: panellists agreed that blind reliance on AI outputs is risky. Clear data provenance, iterative testing, and domain expert validation are essential.
- Ethical frameworks: Alastair Graham called for the creation of shared standards and roles such as model management to ensure responsible governance.
- Market readiness: Alexandra Simmons emphasised the importance of rapid iteration, human-centred testing, and transparency to build confidence among early adopters.
- Skills and education: The panel noted the diminishing need for traditional GIS analysis roles, replaced by new skills in AI interpretation, data stewardship, and ethical oversight.
- Environmental concerns: Alastair Graham advocated for sustainability-first AI infrastructure and transparent energy use reporting.
- Quality assurance: Iris Kramer described ArchAI’s process of iterative model training, error mapping, and manual verification, illustrating that quality is achieved through experience, not automation alone.
Conclusions and next steps
The webinar concluded with reflections on how AI is likely to reshape the geospatial profession over the next decade. While routine geospatial analytical roles may decline, opportunities are expanding for those who combine domain expertise, data literacy, and critical thinking.
All speakers agreed that AI’s value depends on how responsibly it is deployed. Transparency, inclusivity, and ethical awareness are essential to ensure that innovation benefits the broader community—not just the most powerful actors.
Andy Murdock closed the session by encouraging continued collaboration through the RGS SMME Professional Practice Group, inviting participation in upcoming activities and surveys.