We scored 3,029 CVEs for real-world exploitability, here's what we found

For years, discussions about AI and offensive security have run ahead of the evidence. Most claims about what autonomous systems can exploit have come from demos and slide decks rather than data colle