AI SCIENCE
This wasn't in the biology syllabus
Anthropic has revealed that a scientist used its Claude chatbot to help plan potentially risky bird flu experiments.
According to a report released in September, the researcher used Claude to study data, plan experiments and explore changes that could make bird flu spread more easily between mammals.
This kind of research helps scientists understand how viruses could become dangerous to humans.
However, experiments that could make viruses spread faster also raise safety concerns.
In brief:
Experts say there's no evidence the research involved biological weapons or that any experiments took place.
The chatbot may have simply helped the scientist work faster rather than do anything new.
Creating a dangerous virus still requires specialist equipment, trained scientists and hands-on lab work.
A very unfortunate group project
Anthropic discovered the activity in May 2026 and banned the researcher for accessing Claude from a region where it wasn't available.
The company said it couldn't tell whether the researcher had harmful intentions.
Experts also questioned how much Claude helped, as Anthropic shared few details about the case.
I'm beginning to understand why the office printer keeps jamming. It's trying to avoid printing another AI incident report.- MG