ALFA is redefining cybersecurity by uniquely using AI to solve its rapidly emerging challenges. We take an [Adversarial Intelligence and Behavior lens](/unamay) and use clear imperatives to drive our projects. We are driven by the questions defenders fundamentally want to answer. Why was this threat actor doing what has been observed? How does my analysis fit into this threat actor’s plans and goal? What will this threat actor do next? Like puzzle pieces, our projects populate an ecosystem of advances. We have long term investments in projects like [BRON](http://bron.alfa.csail.mit.edu/info.html) which supports threat actor profiling, malware analysis intelligence, and technique-based hunting by unifying critical open cyber threat intelligence resources (CTI) such as ATTACK, CVE and DEFEND into a single, bidirectional graph database, RIVALS which supports modeling cyber scenarios' coevolutionary arms races, and our agent intelligence effort emblemized by our [*LLMs Killed the SCRIPT-KIDDIE* report](https://arxiv.org/abs/2310.06936). We have the acumen to seize an emerging situation. Hence our latest investment in AI-security culture where we are evaluating and comparing organizations' perceptions of AI risks and offering actionable insights into cultural improvements.
0 commit comments