The more information that comes out about the Chinese researcher who edited the genes of two embryos—now twin newborn girls—the more other experts are concerned about the process the scientist followed, the gene he targeted and the results he achieved.

In a 30-minute talk Wednesday at a gene editing summit in Hong Kong, He Jiankui of South University of Science and Technology of China, in Shenzhen, laid out the steps he took to perform the first-ever editing of the human germline—cells that are passed on by reproduction. By the standards of the global scientific community, He used a gene editing technology that is not fully understood, implanted embryos with clear evidence of unusual genetic activity, failed to obtain proper consent from the families involved, and—perhaps worst of all—conducted a risky procedure with no certain benefit.

“The clinical use of genome editing in the germline of humans is premature given that we don’t fully understand the technology and the impact of that kind of application, and it has a potentially very longstanding effect on a person’s life and their future kids,” says Jennifer Doudna, who helped develop the CRISPR-Cas9 gene editing technique He used. Doudna, a professor at the University of California, Berkeley, says she hopes CRISPR will someday be used to edit terrible diseases out of the human genome. But she notes, "one of the dangers of this kind of—I feel premature—study is that it muddies the waters and tarnishes the technology in this area at a time when we should be celebrating it and working together carefully to ensure that it’s used safely.”

The researcher told the audience of several hundred scientists and journalists—with 8,000 more watching live online—that he had edited 30 embryos using CRISPR/Cas9 to delete both copies of the CCR5 gene. CCR5 is naturally absent in about 1 percent of Europeans, leaving them essentially immune to HIV. He said he then implanted two of those embryos into the mother, and the twins were born this fall. He claims all of his tests of the girls showed they were completely healthy and normal.

He, who was trained as a bioengineer (completing his doctorate at Rice University in Houston and a post-doctoral fellowship in a prominent lab at Stanford University) seemed taken aback by the firestorm his work has created. Several scientists, including Doudna, described him as “naive.” He did not respond to a request for comment on his work.

Looking pale and sounding anxious, He told the audience he was proud of his gene editing work. An expert in vaccine development, He said his aim is to rid the world of HIV, the virus that causes AIDS. There is tremendous suffering particularly in the developing world both from the disease and from the stigma surrounding it, he noted several times.

But others say the data suggests cause for concern and question his choice of the CCR5 gene.

The global scientific community decided back in 2015, when CRISPR editing of human genes first became feasible, that gene editing should begin by addressing hereditary diseases, such as sickle cell anemia and cystic fibrosis. Both cause tremendous suffering, are difficult to treat effectively, and in rare cases are certain to be passed to any biological children, says Harvard Medical School Dean George Daley, a stem cell scientist. CCR5 was not a high-priority gene to edit, he says, because there are other ways to effectively prevent and treat HIV.

The girls’ father has HIV, but it is well controlled by medication, so they are at no greater chance of contracting the disease than any other child, says Kiran Musunuru, an associate professor and regenerative biology expert at the University of Pennsylvania. “I think it’s a puzzling choice and a poor choice,” Musunuru says of targeting the CCR5 gene. Plus, lacking CCR5 also makes people more vulnerable to severe reactions or even death from other viruses, potentially including the flu, which is far more common than HIV, Musunuru says. “When you balance the minimal benefit in terms of protection against HIV, compared to increased chance of dying from a fairly common infection like the flu, I don’t think that works out on balance.”

The researcher intended to delete both copies of the CCR5 gene, but one of the twins still has a single copy of the gene, so she will not have any extra protection against HIV.

Even more concerning, Musunuru says, is that tests conducted on the newborns, and when they were embryos, suggest both girls may have mosaicism, which means the edits worked on only some of their cells. In such an unprecedented situation, it is unclear whether the girl with the double deletion will have any protection against HIV, or whether either girl will see any benefit or harm from the gene editing, Musunuru says.

Scientists do not know whether there were any unintended edits to other genes, Musunuru says. Such “off-target” effects are a common result of this form of gene editing and are one of the reasons many scientists believe it is still too early to test the technology on babies.

Editing CCR5 may also have an effect on the brain. Research in mice suggests that deleting CCR5 boosts visual and spatial memory, according to a 2016 study led by Alcino Silva of the University of California, Los Angeles. Silva said via e-mail that deleting both copies of the gene showed greater benefit, but even “one copy alone is sufficient to see enhancements in certain forms of memory.” Other researchers at U.C.L.A., including neurology chairman S. Thomas Carmichael, are currently conducting a clinical trial using the HIV drug Maraviroc to deprive stroke victims of CCR5, hoping it will boost their recovery. If the drug is successful, it would suggest that the human brain can also benefit from deleting the CCR5 gene.

The researcher acknowledged during his talk that he had seen the mouse data on CCR5 but dismissed it. “He rather evaded that question,” says Robin Lovell-Badge, the British stem cell scientist who moderated the discussion with He and asked him on stage about his work, including the potential brain effects of editing CCR5. Lovell-Badge says he does not think He was aiming for genetic enhancement when editing the girls’ genes, however.

Daley, for his part, says he was surprised to hear that He knew about the brain research. “I can’t know his motives, I just think it’s curious that he knew about the paper.”

On his website and in a video posted to YouTube, He argues that gene editing is only ethical for the purpose of healing. He says it is totally inappropriate to use gene editing to select hair or eye color, or to boost IQ; such a use, he says in the video, “is not what a loving parent does. That should be banned.”

It is not clear what will happen now to He, who is being investigated both by Chinese authorities and his home university. He has been on leave from the university since February, and the school said in a statement it was unaware of his work in humans and condemned it. He says there is currently another early pregnancy that resulted from his gene-edited embryos.

Organizers of the Hong Kong summit, including Doudna, Lovell-Badge and Daley, are meeting Thursday to decide whether to issue a formal response to He’s gene editing work.