Can some human groups evolve into new species?

What would it actually take for a human group to become a new species?

Speciation means a population splits into groups that no longer interbreed and then diverge enough to count as separate species. Modern speciation theory says you need to identify the ancestral population and its preexisting traits, the evolutionary forces that drive the main differences, and the salient adaptations that mark the new species [3]. In ordinary terms: a group has to be cut off from the rest, face different pressures, and accumulate enough change that it can no longer merge back.

That is not a quick or automatic process. The same framework applies to humans as to other animals, which means the question is less about whether humans are special and more about whether any human group could become isolated enough, for long enough, for divergence to stick [3]. The theory describes the conditions; it does not claim that any current human group is on that path.

Humans have already done this before

The strongest evidence that human groups can become new species is that they already have. Direct dating of human fossils has helped identify and date several newly named species, including Homo naledi, Homo luzonensis, and Homo longi [4]. These are not hypothetical branches; they are named species recovered from the fossil record and placed in time through geochronological analysis.

Dating has also rewritten specific stories. Homo floresiensis was redated to more than 60,000 years ago rather than about 18,000 years ago, which ended a heated debate about how it could have survived alongside Homo sapiens on Flores for tens of thousands of years [4]. The takeaway for the reader is that human evolution has produced multiple coexisting species, not a single line. Whether any living human group is currently splitting is a different question from whether the human family has split in the past.

Could technology or AI produce a new human species?

Some technocratic thinkers assume human evolution will soon lead to a new species of hominids. Ray Kurzweil's 'Singularity' hypothesis imagines a fusion of humans with software carriers and minds with CPU capabilities, while Yuval Noah Harari's 'homo deus' envisions a next step armed with artificial intelligence [1]. A soberer alternative, Wolfgang Sassin's 'homo billionis,' suggests a next generation adopting herd behaviour under uncontrolled globalisation [1].

These are predictions and ideological positions, not demonstrated evolutionary outcomes. The review notes that Vaclav Smil treats growth as an evolutionary process but overlooks the reasons and mechanisms of evolution, and that he proposes no instruments for assessing AI's involvement in human growth [1]. So the honest reading is that a technologically driven new human species is a forecast some thinkers make, not something the evidence shows is happening.

Why some researchers reject the whole framework

Not everyone accepts the standard account. One line of argument claims that Neo-Darwinism and population genetics are incompatible with Hardy-Weinberg's Law and Mendel's Laws, that mutations are harmful rather than creative, and therefore that no evolution occurs at all [2]. A related argument holds that hybridization cannot produce a fertile, reproductively isolated offspring, so new species do not arise artificially or naturally [5].

These are minority critiques, and they are presented as arguments against evolutionary mechanisms rather than as evidence of human speciation. For a general reader, the practical point is that the mainstream case for human groups having speciated rests on the fossil and dating record [4] and on speciation theory [3], while the rejectionist case disputes the underlying mechanisms [2][5]. The two positions do not agree on what counts as evidence, so the disagreement is about the framework itself, not about a single disputed fossil.

Sources used in this answer

Related research questions