Maps, trust and AI: the long history of cartography’s accuracy and its modern challenges

Overview

The Conversation examines how centuries of mapmaking built trust in geographic accuracy, how myths like the Kong Mountains persisted across maps, and how modern AI features in mapping are testing that bedrock of trust. It traces the arc from early courts and publishers to the digital era where AI can conjure convincing but false satellite images, prompting scrutiny from fact-checkers and researchers alike.

• Historical trust in map accuracy shaped industrialisation and colonial perspectives.
• The Kong Mountains myth shows how cartographers extended unseen features, sometimes for political or practical reasons.
• Early copyright rulings allowed improvements and warned against slavish imitation, a theme echoed in AI data use today.
• Google Earth’s AI features sparked rapid safeguards and public debate on trust, accuracy, and the need for credible sources.

Author: The Conversation

Overview

This piece traces how Europe’s cartographers built a reputation for precision during a time of vast social change, and how demand for ever more detailed world maps during industrialisation and colonisation sometimes outpaced accuracy. It uses a historical through-line to illustrate how maps shape perception and how trust in the mapping process has been both crucial and fragile.

The Kong Mountains and the expansion of a map’s world

The article recounts the Mountains of Kong, which appeared on African maps after Mungo Park described them to his cartographer James Rennell in the 1790s. Rennell produced two maps featuring this mountain chain and extended it far beyond Park’s route. The mountains were widely copied, and doubts about their existence persisted into the late 19th century. This episode reveals how a cartographer’s trustworthiness could propagate—true or not—across widely disseminated materials, shaping how people understood geography.

The turning point and the persistence of myth

In 1887–1888 a French officer Louis-Gustave Binger undertook an expedition that proved Kong Mountains did not exist. Yet references lingered; the author notes finding a 1936 Oxford Advanced Atlas index reference to Kong Mountains in The Oxford Advanced Atlas, nearly 50 years after Binger’s expedition. The episode underlines how maps can imprint a worldview even after contradictions are established, highlighting the lasting impact of cartographic authority.

From maps to copyright and AI

The article connects historical mapmaking to modern questions about copyright protections that are being tested by AI today. It cites a London publisher Robert Sayer who sued his rival John Hamilton Moore in 1785 for copying nautical charts; the court ruled in Moore’s favour, establishing that earlier charts could be consulted and improved as long as the new work was not merely a servile imitation but an improvement. This ruling has been invoked in discussions about AI’s use of copyrighted material to generate new content, suggesting a precedent for balancing transformation with fairness.

Evidence, trust and the AI era

The narrative then shifts to how trust in imaging technology has evolved from ancient instruments to satellites. Google added generative AI to its Google Earth platform, enabling the creation of convincing but fictional satellite imagery. This sparked public outcry and swift action: Google rolled back the AI tool within 24 hours and pledged stronger guardrails. Fact-checkers like BBC Verify weighed in as trusted intermediaries in identifying misinformation.

The piece also notes how investigative groups such as Bellingcat used verification methods to expose deepfakes in other contexts, such as analyzing satellite imagery surrounding Malaysia Airlines MH17. The author reflects on a line spoken by a former Google geospatial technologist: maps were once described as a mirror of the real world, but that trust was challenged by AI-generated content.

Lessons for the future

Across centuries, the article argues, trust in maps has been the bedrock of how we view the world. In the current era, where AI can reconstitute copyrighted material in non‑imitative ways, it is vital to preserve credible, expertly vetted summaries and cross-linkings. The piece closes with a reminder that the core values of mapmaking—accuracy, transparency and responsibility—remain essential as technology enables new frontiers in data and visualization.

Author: The Conversation