Applied Technology Lab

A new research lab at Harvard

We build technology in the public interest.

Get in touch

When AI goes rogue

"AI safety risks are not a theoretical problem. They haven’t been for quite some time. And it’s naive for people to think that we can afford to not think about safety risks because we want these AIs to solve cancer."

3 min read

What we’re working on

Encrypted Spaces

Privacy is normal.

Open-source architecture for end-to-end encrypted collaborative applications.

Keyring

Take back your identity.

Portable, interoperable, privacy-preserving identity infrastructure.

Guillotine

AI containment is a systems problem.

A hypervisor architecture for isolating malicious or misbehaving AI across software, hardware, networks, and physical infrastructure.

Transparency Hub

Tracking tech company policy changes over time.

Archiving policy documents to enable research and encourage informed consumer consent.

Our goal is research with impact

We make software that developers, researchers, and organizations can use, inspect, adapt, and improve.

How we choose our projects

  1. The risk is near-term.
  2. The technical gap is real.
  3. A small, highly skilled team can make meaningful progress.
  4. The outputs are concrete and useful.

Our focus

Right now the lab is focusing on AI systems and agents, future-proof cybersecurity, digital identity, and applied cryptography.

Who we are

Partner with us.

If you're interested in partnering with the lab, please reach out using this form.

Get in touch