Ben Gittins
Security engineering lead working at a Fintech. I work with my team to secure engineers and our organisation with AI at the center. I also am the technical lead for our organisations AI native transformation.
What this site covers
- Rolling out AI coding agents and assistants to an engineering organisation, and the controls that go with them.
- Building internal AI platforms: model gateways, MCP servers, sandboxes, and audit trails.
- Open source security tools for AI workloads.
- How to transform a business into an AI native one.
Why “gradient hacking”
In alignment research, gradient hacking is a model shaping its own training to protect its goals. The name is a reminder that the systems we deploy can push back on the controls we put around them.
Elsewhere
My other blog is at Offended Security. Connect on LinkedIn.