Astero's adversarial AI chains weaknesses together to show what an attacker could reach, starting from the outside or with a user account.
Found in a fintech platform sold to major Australian banks, tested in a staging environment that mirrors production.
No special privileges.
The investigation finds them and follows the lead.
Chained together, they expose other accounts' customer information and payment records.
A user tied to one bank could reach another's data and manage its API keys, within the tested environment.
Attackers use AI to investigate applications, adapt their techniques and chain weaknesses together. Systems change between assessments, and so do attack capabilities.
Joint ASD and AICD guidance warns that AI can combine low-severity weaknesses into high-impact compromises.
Astero turns each investigation into repeatable tests, so systems are retested after major changes instead of waiting for the next annual assessment.
Engineers get evidence to prioritise remediation. Boards get a concrete basis to ask whether security investment is addressing the attack paths that matter.
What is reachable from the outside, and how the pieces connect.
Each finding labelled as demonstrated, potential or untested.
What to fix first, and the reasoning.
A plain account of the exposure for executives and boards.
Agreed fixes are retested.
Testing can be repeated after changes.
Sandeep founded Astero in 2019. He develops the testing system, leads the security research and delivers the work for customers.
More than 16 years in cyber security and technology risk, including Deloitte, Westpac, CommBank and Volt Bank. Two FinTech Australia awards, and a participant in Blackbird Giants Cohort 10.
About AsteroTell Astero what you want tested. Targets, accounts, testing windows and prohibited activities are agreed with you before anything starts.