Project summary

We find it more effective to develop technical solutions rather than talk about the risks posed by artificial intelligence. To solve this problem, we are founding SolonGate Research which focuses on AI safety and human control.

And now we want to take it further with a small Research fellowship, giving young engineers the opportunity to work on meaningful AI safety problems and share what they discover with everyone.

Good research shouldn't require a famous university badge or million dollar lab :)

https://solongate.org/fellowship

What are this project's goals? How will you achieve them?

We'll support one or two researchers for eight weeks, working on AI safety. ( it can be everything in this field) Research directions include safety evaluations, human oversight, autonomous systems, privacy, and open security infrastructure. We will help fellows turn specific research questions into experiments like open-source tools, or technical reports. (The results will be public, including the failures)

In the end we will release all the outputs from the process as open source projects. Tangible technical outputs are very important to us.

How will you know it worked?

• Our primary success criterion is for researchers to publish concrete technical work. Subsequently, applying these technical studies in real world scenarios will fulfill our mission
  

How will this funding be used?

• the funding covers one fellow's stipend ( 2400 )

• Mentor and engineering support (1800)

• technical review (700)

• infrastructure (600)

• publication ( 500 )
  

Who is on your team? What's your track record on similar projects?

We're Emirhan and Çınar, two 19 yearold founders of SolonGate.

We've open sourced our security tools and started receiving contributions from developers around the world. And also our software has crossed 450k npm downloads.

It's honestly pretty surreal seeing people on the other side of the planet contributing to something we started :)

Our open-source projects

• AI agent Security Gateway: https://github.com/solongate/agent-security

• PSIRTMap: https://github.com/solongate/psirtmap

Our research & benchmarks

• Agent Security Benchmark: https://solongate.com/blog/agent-guardrail-benchmark/

• Shadow AI Leak Benchmark: https://solongate.com/blog/shadow-ai-leak-benchmark/

We're not a big research institution or something like that. As two 19 yearold researchers, we are sharing our work in this field in a completely open manner. Our goal is to transform this into a collective movement...

What are the most likely causes and outcomes if this project fails?

There is no possibility of failure, because regardless of whether the results of the research and studies are negative or positive, all outcomes will be translated into concrete reports and open-source projects.

How much money have you raised in the last 12 months, and from where?

We haven't received dedicated funding for SolonGate Research or the fellowship.

https://solongate.org

Just look :)