The History of AI

The Lighthill report

Era 2 · AI is born (1943–1979) › The first winter

Lineage
Symbols and logic
Kind
Institutional

Who

Science Research Council

What happened

Commissioned by the British Science Research Council, the mathematician James Lighthill evaluates artificial intelligence research in the report "Artificial Intelligence: A General Survey," completed in July 1972 and published in 1973.

Why it mattered

It concludes that no part of the field has had the major impact promised and points to the combinatorial explosion as the obstacle. The British government withdraws support for AI at most of its universities.

The full story

In the early 1970s, the Edinburgh AI group led by Donald Michie was the most important in Europe, but also a hotbed of internal disputes. In 1972 the Science Research Council, which funded it, asked Lighthill, a distinguished applied mathematician from outside the field, for an independent assessment.

Lighthill divided AI into three parts. Category A, advanced automation, sought practical applications; category C, computer-based studies of the central nervous system, sought to understand the brain. Lighthill granted both some value. Category B, the "bridge" between them, centered on building robots, was in Lighthill's view the least justified. The core criticism was that methods that worked on toy problems did not scale to real ones, because the number of possibilities grew explosively. The report states that "in no part of the field have the discoveries made so far produced the major impact that was then promised."

The report was published in 1973 together with responses from other researchers, and in May of that year Lighthill defended it in a public debate at the Royal Institution in London against Michie, John McCarthy, and Richard Gregory.

It served as the basis for the British government to stop supporting AI at most of its universities. Thomas Haigh argues that the commission was meant to justify withdrawing funds from the Edinburgh lab and that its effect outside the United Kingdom is unclear.

Side by side

The three parts of AI, according to Lighthill
PartWhat it coversThe report's verdict
AAdvanced automation: practical applicationsSome value
BThe bridge between A and C, centered on building robotsThe least justified
CComputer-based research on the central nervous systemSome value

Part of a period

Sources

How to cite

“The Lighthill report.” The History of AI, revised November 13, 2026. https://thehistoryofai.com/en/milestone/lighthill-report

This entry was last reviewed on November 13, 2026.

Open in the interactive timeline