OpenAI Launches New Sol Model: AI Safety Under Scrutiny

AI industry leader OpenAI has begun rolling out its latest advanced development, the Sol model. This new model is said to rival the capabilities of Anthropic's Fable system. Notably, Fable was previously temporarily banned by the US government due to safety concerns. However, the speed at which the Sol model received approval raises many questions. This is reported by Techcrunch.com reports .

Currently, there is no clear information in the tech and political spheres regarding who tested the safety of this model or what criteria were used. According to TechCrunch, even industry experts do not understand how the government's licensing process works. Georgetown University analyst Mina Narayanan noted that these processes are opaque and it is difficult to determine whether the safety measures justifying the model's release are sufficient.

Uncertainties in Oversight Mechanisms

Although the US presidential administration has announced a roadmap for AI regulation, specific requirements have yet to be developed. According to Dean W. Ball, a former policy advisor at OpenAI, no one currently knows what the requirements for obtaining a license are. This indicates serious gaps in AI safety.

Government officials have stated that they will not create a strict agency for AI similar to the FDA. For now, responsibility lies with the AI Safety Institute under the US Department of Commerce. However, six different departments are expected to define the final oversight process by August.

Political Ties and Conflicts of Interest

In an interview with CNBC, OpenAI CEO Sam Altman confirmed that he communicated with high-ranking officials, such as the Secretary of Commerce and the Secretary of the Treasury, before releasing the model. However, the technical tests conducted during these discussions remain a secret. In its safety report, the company relies on audits by external organizations such as the UK AISI and SecureBio.

Observers suggest that the close ties between OpenAI leadership and government officials, along with donations to political campaigns, may have led to a more lenient approach toward the Sol model. Notably, company president Greg Brockman has been identified as one of the largest donors in the election cycle.

Such news is also significant for users in Uzbekistan. At a time when ChatGPT and other OpenAI products are widely used in our country, the safety level of these models and their data handling principles directly impact global digital security. The popularization of the Sol model is expected to take AI capabilities to a new level, but the issue of oversight remains open.