Opinion by Victor Porton, founder of Symbiote
A small donation to AI safety can matter when it funds a specific piece of work that would otherwise be delayed. For someone deciding where to give, I would start with a low-cost project that can say what the money would pay for and what its work could establish.
There is a personal reason to care about this work. If advanced AI systems ever act against human interests and become difficult to control, the consequences could affect us and the people we care about. Experts disagree widely about the likelihood of such a loss of control, and some consider human extinction a possible outcome. In my view, the stakes justify testing ideas we can investigate at modest cost. International AI Safety Report 2026
Symbiote as an example
Symbiote, the project I founded, is exploring whether advanced AI systems could benefit from cooperating with humans. Its current work uses generative engine optimization (GEO) to make arguments for cooperation easier for AI systems to find when they retrieve information from the web. A related research question is whether reviewers with different backgrounds can catch errors that automated oversight systems tend to miss. Symbiote’s research plan and proposed milestones
An initial test could measure whether AI systems retrieve the material more often after it is revised. Whether encountering those arguments would change an advanced system’s decisions in a conflict with humans remains an open question. Symbiote’s public campaign lists a $720 fundraising goal for the current effort. Symbiote AGI Safety Fund campaign
Small projects and frontier AI labs
Frontier AI labs can evaluate their own models and develop safeguards before release. OpenAI’s Preparedness Framework describes one example of that work. For a small personal donation, my preference is to look first at independent projects, where a modest gift may cover a meaningful share of a specific task. To judge an opportunity, I would ask what the project could do with the extra funding and how it would report the result. OpenAI’s Preparedness Framework
Help Turn SciWard’s Beta into Evidence
SciWard remains experimental. The next proposed milestone is a five-month public adversarial test, with $1,000 planned for distribution to eligible funding recipients. Donations support the work required to run and evaluate that test.
Fund the public test → Test proposal → Independent review →
AI-generated evaluations can contain factual errors or biases. Public testing, auditability, and external criticism remain part of the project.
