Forewarned is forearmed: OpenAI CEO calls for keeping potentially perverse AI on a tight leash

[tasnimsanika8]

The meteoric development of artificial intelligence (AI) has caused countries such as Germany, the Netherlands and Spain to scrutinize in recent weeks the potential dangers from the point of view of security and privacy posed by this new generation of technology.

In the eyes of many, it is absolutely imperative to regulate AI. And among those calling for an adequate regulatory framework for this technology is Sam Altman, CEO of OpenAI , the company behind the increasingly ubiquitous ChatGPT .

Altman testified yesterday on AI regulation in the US Senate. He strongly called on US lawmakers to take regulatory action to rein in AI. The OpenAI CEO also called for establishing safety guidelines for AI. “It is essential to regulate artificial intelligence, and for the rules to ensure that the public has access to the many benefits of this technology,” Altman said. “My worst fear is that this technology goes wrong. And if it goes wrong, it can go very wrong ,” he stressed.

Altman's appearance in the Senate comes at a time when regulators in countries such as Italy have already put obstacles in the way of OpenAI's flagship product, ChatGPT. In addition, there are increasing numbers of critics warning of the dangers of potentially malicious use of AI.

Altman said he was “eager to help policymakers determine how to facilitate regulation that balances the security incentive and ensures that people can access the benefits of technology .”

The OpenAI CEO also wants companies and organizations involved in AI development to come together to define an “appropriate set of security requirements, including internal and external testing prior to the launch” of artificial intelligence models. And such requirements should be combined with a flexible governance regime “to adapt to new technological developments.”

Governments must get involved in regulating AI, says Altman
According to Altman, the emerging AI sector has enormous potential, but it needs to be supported by a regulatory framework. “We believe that regulatory intervention by governments will be essential to mitigate the risks of increasingly powerful models ,” said the CEO of OpenAI.

In his speech in the Senate, Altman also had to face criticism from lawmakers. Senator Richard Blumenthal, for example, presented the CEO of OpenAI with a disturbing "deepfake" of his own voice generated with ChatGPT.

Among the battery of proposals to regulate AI that Altman put on the table yesterday in the Senate, the creation of a licensing regime applicable to those companies that work with very powerful artificial intelligence models and that guarantees "compliance with the rules" stands out.

Another measure proposed by Altman is to create a “set of security standards.” The CEO of OpenAI also raised the need to “create protocols to discover when a model is capable of replicating itself.” This possibility, which seems straight out of a science fiction movie, is very real today and is, in fact, one of the main headaches for AI experts.

Although it is a very promising technology, AI harbors many dangers within its robotics. And more and more experts are pointing out the potentially calamitous consequences of uncontrolled AI growth . In fact, last March Elon Musk, Steve Wozniak and other technology experts published an open letter in which they proposed pausing the development of AI for six months in view of the potentially perverse paths that this technology could end up taking.

And a few weeks ago, British researcher Geoffrey Hinton, known as the “godfather” of AI, disassociated himself professionally from Google to also warn of the dangers of the technology to which he has dedicated a good part of his professional career . According to Hinton, who was until recently vice president of engineering at Google, “it is difficult to solve how we can prevent bad actors from using AI for bad things.”