Should Open-Source AI be Regulated?

It’s impossible to escape talking about AI. From everyday usage with ChatGPT to  more evolved machine learning personas in your favorite software, there’s no denying AI’s moment in the sun.  While the applications and benefits for society are numerous, such as improving health care, education, transportation, and entertainment it also poses ethical conundrums and risks. In this blog, we’ll explore both sides of the coin. 

The Case for Regulating AI

AI, like children in school, needs to be taught. A major concern is that the data samples used to train AI models may be biased and lead to  unfair or inaccurate outcomes that affect people’s lives, such as denying them loans, jobs, or health care based on their race, gender, or other characteristics.

Some AI systems may also collect, store, or use personal data without people’s consent or knowledge, exposing them to identity theft, fraud, or surveillance. Theoretically, AI systems may be vulnerable to hacking, manipulation, or sabotage, which could cause physical or digital damage, such as crashing cars, spreading misinformation, or disrupting infrastructure. Therefore, regulating AI can help ensure that AI systems are designed, developed, and deployed in a way that respects human rights, values, and laws, and that they are accountable, transparent, and auditable1

A positive of AI regulation is that it can promote and protect the public interest and welfare, especially in areas that have significant social, economic, or environmental implications. For example, some AI applications may have positive or negative impacts on public health, safety, security, education, or employment. It can also create new opportunities or challenges for innovation, competition, or collaboration.

The Case for Open-Source AI

Innovation and creativity are the two main drivers for open-source AI. The ability to access, modify, and share AI code and resources can foster new developments and use cases.  Open-source AI also facilitates collaboration and cooperation among different individuals and organizations, who can exchange ideas, feedback, and data, and contribute to the improvement and advancement of AI. Eliminating barriers to entry increases participation in the AI field, and can provide more intersectional applications. 

Another argument for open-source AI is that it can enhance and ensure transparency and accountability, by allowing anyone to inspect, verify, and challenge AI code and results. For example, open-source AI can help empower users by revealing the logic, assumptions, and data behind AI decisions and actions, and expose any errors, flaws, or biases that may affect their quality or reliability. It can also help monitor and audit AI systems and processes, and hold them responsible for their outcomes and impacts. 

In Conclusion 

To sum it up, open-source AI and AI regulation are not mutually exclusive, but rather complementary and interdependent. On the one hand, open-source AI can support and facilitate AI regulation, by providing more information, evidence, and feedback for regulators and policymakers, and by enabling more participation, consultation, and collaboration among different actors and interests. On the other hand, AI regulation can support and facilitate open-source AI, by providing more guidance, standards, and incentives for developers and users, and by ensuring more protection, fairness, and trust for society and the environment. Therefore, both open-source AI and AI regulation are necessary and beneficial for the development and use of AI, and that they should be pursued and implemented in a balanced and coordinated way.