AI Firms' Strict Policies 'Chill' Independent Research on Technology, New Study Says

AI companies' strict policies to prevent bad actors from using their technologies might also be hindering the transparency and trustworthiness of AI systems, independent researchers claim.

In an open letter signed by over 200 digital experts and researchers, the group is calling for AI firms like OpenAI, Google, and Microsoft to provide a "safe harbor" to auditors to explore their technology.

According to them, the companies' terms of service not only deter malicious use of their products but also prevent "independent good faith research" from safety-testing their AI models.

Researchers fear that the companies can suspend their accounts or, even worse, sue them for evaluating their products. The letter accused these policies of causing "chilling effects on research."

AI Development,Assessment Still Locked Behind Closed Doors

Prominent AI companies like OpenAI and Microsoft kept most development and safety tests of their AI models closed doors with evaluation reports coming from their own research teams.

The open letter claimed that access programs for independent researchers promote public awareness and accountability for these AI firms on potential dangers and vulnerabilities their models may have.

Among the signees of the letter were Pulitzer Prize laureate Julia Angwin, Hugging Face CEO Clem Delangue, and Mozilla fellow Deb Raji, who started research auditing AI development.

More Demands for Open AI Probe

The open letter is the latest petition for AI firms to disclose more information on their development amid concerns about the risks posed by the emerging technology.

Just last January, the Commerce Department was reported of leveraging the Defense Product Act to compel tech companies to disclose development plans for their future AI projects.

The state intervention also intends to make the companies report every time a "non-US entity" used their cloud services to train AI models.

The recent calls for a more open dialogue between AI companies and the public reflect an increasing need for regulation of the technology amid its rapid development.