In the EU AI Act, systemic risk means the particular danger very capable AI models can pose to society, safety or fundamental rights.
Example
A very large model that could help with cyberattacks falls under stricter obligations.
For pros
Under Art. 51, a GPAI model qualifies if it has high-impact capabilities; this is presumed above 10^25 floating-point operations (FLOP) of training compute, or the Commission designates it. Providers must then, among other things, assess risks, run adversarial tests, report serious incidents and ensure cybersecurity (Art. 55).
General information only, not legal advice. For your specific case, ask a qualified lawyer.
