High-containment laboratories are expanding rapidly around the world.
Strong, consistent safety standards are essential to prevent laboratory accidents and potential pandemics.
Update: Bhattacharya asked Nature Medicine to reply within 30 days (around Dec 18) with any plan for a review. Co-author Ian Lipkin told Science the paper "definitely was not ghostwritten."
science.org/content/article/…
NIH Director Jay Bhattacharya asked Nature Medicine to investigate whether Fauci, Collins and Jeremy Farrar made uncredited contributions to the 2020 "Proximal Origin" paper that dismissed a COVID lab origin, and to correct or retract it if warranted.
politico.com/news/2026/11/18…
NIH has named John H. Powers III permanent director of NIAID. NIH Director Jay Bhattacharya told the NY Post: “No more dangerous gain of function will happen” under the new leadership, and NIAID is “now purely a civilian agency.” nih.gov/news-events/news-rel…
With Democrats taking the Senate, Sen. Rand Paul, the chief congressional driver of gain-of-function oversight at HSGAC, is weighing a bid for top Republican on the HELP Committee. Where biosafety oversight lands next matters. washingtonexaminer.com/news/…
White House released the National Policy for Nucleic Acid Synthesis Screening (NSTM-6, Nov 5). Federally funded labs must procure synthetic DNA (≥50 nt) and benchtop synthesizers only from certified suppliers using NIST-verified screening—replacing the 2024 self-attestation framework. Explicitly supports the Stopping Policy on dangerous GoF.
whitehouse.gov/wp-content/up…
1
8
353
A press release is not oversight.
If CREID is back, the grant files and the IBC reviews on those awards have to be public, and a violation has to cost the award—not a statement about “highest possible standards.”
Replying to @NIH_Innovates @NIHDirector_Jay
Goal 5 is the part that lasts — invest in the boxes, not just the experiments. Collaborative oversight among maintenance, safety officers, and end users is the right order.
20
Replying to @Bryce_Nickels
At one point in the next in series, can you ask who signs off on AI releases?
24
Replying to @Bryce_Nickels @R_H_Ebright
Two people. One page. Both have to sign before the model ships. One knows pathogens. One can halt the release.
28
Replying to @R_H_Ebright
The gap is the same one that failed before 9/11. Virologists don’t sit on the call when the model ships. Shutdown authority has to live inside the lab, with a virologist who knows pathogens on the line — not a memo after release.
1
24
Replying to @R_H_Ebright
The people with the power to slow the next GPT release sit on OpenAI’s Safety and Security Committee — Zico Kolter chairs it, Paul Christiano just joined. They don’t have a virologist in that room. You should be the one they call.
1
3
51
Replying to @Bryce_Nickels
If there’s a question slot Friday — whether the AI labs and the virologists share a room yet.
1
25
Replying to @hilbertspaess
The people who built it are walking out. The people who would have to contain a pathogen never got a seat. That’s the gap.
1
23
Replying to @michaelshermer @robertwrighter
The AI labs and the virologists still don’t share a room. Nobody in the room sees both sides. @Bryce_Nickels
22
Replying to @Bryce_Nickels @RutgersU
Fair. If it’s not a person, it’s a function — someone has to sit between the two rooms before the next model ships. Simon’s point is the gap. That’s the argument.
31
Bryce — do you or Richard know Kevin Esvelt well enough to know if he’d be the right person for the triage role? The guy who understands both rooms. The screenshots I just posted under your announcement are the argument. Jeff Bell here.
1
42
Speaking up is the duty — but the people who could speak up are split across two rooms that never meet. The AI labs don’t understand virology, the virologists don’t understand AI. Same stovepiping that failed before 9/11. Nobody in the room sees both sides.
2
28






