OpenAI acknowledged the increased risk of producing biological weapons with the o1 model

OpenAI’s latest artificial intelligence model faces an increased risk of being misused to make biological weapons, and security researchers are sounding the alarm over the risks of this emerging technology. Interestingly, OpenAI itself is aware of these risks and has acknowledged it.

According to the Financial Times, the company OpenAI introduced its new model called o1 on Thursday, announcing new capabilities for reasoning, solving difficult mathematical problems and answering scientific research questions. These abilities are seen as a critical advance for creating comprehensive artificial intelligence or machines with human-level cognition.

OpenAI says the new models pose a moderate risk for chemical, biological, radiological and nuclear (CBRN) weapons issues, in part of its how-to toolkit.

The “medium” level is the highest category OpenAI has ever given to its models; That is, the technology “significantly provides assistance that enhances the ability of professionals in advanced CBRN-related fields to develop a CBRN threat.”

With more advanced capabilities, such as the ability to run step-by-step reasoning, this artificial intelligence software increases the risk of abuse by bad actors, experts say.

These warnings are raised while technology companies such as Google, Meta and Entropic are competing to build and improve advanced artificial intelligence systems; Because they are looking to create software that can act as “agents” that will help people complete their tasks and lead their lives.

California’s SB 1047 bill and OpenAI’s opposition to it

This has led to efforts to legislate AI companies. In California, debate is raging over a bill called SB 1047 that would require creators of artificial intelligence models to take steps to minimize the risk of their models being used to develop biological weapons.

Some venture capitalists and technology groups, including OpenAI, have warned that the proposed rule could have a devastating effect on the AI ​​industry. Gavin Newsom, the governor of California, must decide in the coming days whether to sign the law or veto it.