Penn launches AI cybersecurity centre with former CISA and NSA officials

The University of Pennsylvania has established an engineering research centre dedicated to the intersection of cybersecurity and artificial intelligence.

Based within Penn’s School of Engineering and Applied Science, the Schlein Center will examine defensive computing architectures and automated digital risks. The initiative is backed by a founding commitment from venture capitalist Ted Schlein. It aims to accelerate technical counter-measures against autonomous vulnerabilities.

To guide its operational focus, the centre recruited former senior figures from the Cybersecurity and Infrastructure Security Agency and the National Security Agency. By integrating veteran intelligence officials, Penn intends to bridge the divide between theoretical computer science and active network defence. These leaders will help direct research toward vulnerabilities that threaten government systems and civilian infrastructure.

Faculty and research fellows will study the dual-use profile of frontier machine learning models. Early technical priorities include auditing how autonomous agents discover zero-day flaws, countering automated exploit generation, and securing critical digital utilities. Researchers also plan to investigate adversarial inputs intended to corrupt machine learning models and manipulate network telemetry.

Penn Engineering designed the centre to operate as a cross-disciplinary engineering hub. The project links computer science specialists with electrical engineers to build tamper-resistant hardware and verified software protocols. Alongside standard peer-reviewed literature, the facility plans to publish open benchmarks and defensive toolkits for external security teams.

The launch reflects an accelerating pivot across technical universities. Academic institutions increasingly treat algorithmic safety, systems engineering, and national security as unified operational domains. As automated scanning reduces the friction of discovering vulnerabilities, academic labs face pressure to supply verifiable defences to the public sphere.

Can university laboratories produce automated defensive systems quickly enough to neutralise state-tier offensive tooling before such tools enter wide circulation?