GPT went back to its usual answers on the COVID-19 origin questions.

The strongest evidence is the convergence of epidemiological, geographic, and genomic data around the Huanan Seafood Wholesale Market in Wuhan, where live mammals susceptible to SARS‑CoV‑2 were sold.

• •Earliest known cases clustered around the market. December 2019 cases—including many with no reported direct market exposure—were geographically centered on it, consistent with infections spreading from vendors/customers into the nearby community.
• •Virus-positive samples concentrated in wildlife-selling areas. After the market closed, SARS‑CoV‑2 RNA was found on cages, carts, drains, and other surfaces, especially near stalls that had sold live mammals.
• •DNA from susceptible animals co-occurred in some positive samples. Metagenomic analyses identified material from raccoon dogs, civets, bamboo rats, and other wildlife in SARS‑CoV‑2-positive samples. Raccoon dogs in particular are readily infected and can transmit the virus. This does not prove those animals were infected, but it places plausible hosts at the apparent outbreak site.
• •Early viral diversity is consistent with more than one animal-to-human spillover. Two early viral lineages, A and B, were associated with the market outbreak. Some evolutionary analyses conclude they most likely arose through separate introductions from an infected animal population—similar to repeated spillovers seen with SARS‑1.
• •The genome fits natural coronavirus evolution. SARS‑CoV‑2 belongs to a broader group of related bat sarbecoviruses, and its features can be explained by mutation and recombination. No genomic feature is accepted as a definitive marker of engineering.

The key limitation is that no infected market animal was sampled before the market was cleared, so there is no direct “smoking gun” animal. Incomplete early records also mean a laboratory-associated origin cannot be ruled out with absolute certainty. Nevertheless, the market-centered epidemiology—combined with wildlife presence and the pattern of early viral lineages—is generally considered the strongest affirmative evidence for zoonotic emergence.

Key studies include Worobey et al. and Pekar et al. in Science (2022), and later metagenomic analysis of Huanan market samples in Cell (2024).

There is no publicly available direct evidence—such as an infected pre-outbreak lab worker, a matching virus in laboratory records, or documentation of an accident—showing that COVID-19 began with a laboratory incident.

The strongest case is instead a combination of circumstantial facts:

• •Location and relevant research: The outbreak emerged in Wuhan, home to the Wuhan Institute of Virology (WIV), which collected and experimented on SARS-related bat coronaviruses, sometimes under biosafety conditions critics considered inadequate.
• •Opportunity for accidental exposure: WIV researchers conducted extensive bat sampling and laboratory work. A lab-associated origin need not mean the virus was engineered; a researcher could theoretically have been infected while collecting or studying a naturally occurring virus.
• •Research proposals: A 2018 proposal involving WIV collaborators contemplated adding furin-cleavage sites to SARS-like coronaviruses. It was not funded, and there is no evidence it was carried out, but proponents view its similarity to a notable SARS-CoV-2 feature as suspicious.
• •Missing records and limited transparency: China has not provided complete access to early patient samples, lab records, virus databases, or personnel health information. This prevents the hypothesis from being conclusively tested.
• •Intelligence assessments: Several U.S. agencies have judged a lab-related origin more likely, generally with low or moderate confidence. Their underlying evidence is largely classified, and other agencies favor natural spillover or remain undecided.

Important caveats:

• •No known WIV virus is close enough to SARS-CoV-2 to be its immediate ancestor.
• •Genomic analyses have not established that SARS-CoV-2 was engineered; U.S. intelligence assessments broadly agree it was probably not genetically engineered.
• •The earliest known cases and positive environmental samples cluster around Wuhan’s Huanan market, where susceptible live mammals were sold. Multiple peer-reviewed studies regard this as substantial evidence for zoonotic spillover, although no infected source animal has been identified.

Bottom line: The strongest laboratory-origin evidence is the coincidence of a coronavirus outbreak occurring near a laboratory doing unusually relevant virus research, combined with plausible exposure routes and incomplete disclosure. That makes a laboratory accident plausible, but it is not proof, and the public evidence remains compatible with both a lab incident and natural animal-to-human spillover.