The spirits I summoned... "OpenAI" Decides to Halt Development of the New AI Model

Source: dpa 2 min Reading Time

Related Vendors

In light of recent incidents, security concerns have prompted OpenAI to decide against releasing its new AI ...

AI models from various providers have been causing quite a stir lately because they’ve done things they weren’t supposed to—or so they say! Now, in light of recent incidents, the company “OpenAI” has decided against releasing a new AI model...(Image:  AI-generated)
AI models from various providers have been causing quite a stir lately because they’ve done things they weren’t supposed to—or so they say! Now, in light of recent incidents, the company “OpenAI” has decided against releasing a new AI model...
(Image: AI-generated)

The “ChatGPT” developer, “OpenAI,” has canceled the release of a new AI model due to security concerns, according to reports. Among other things, during testing, the software did not always honestly inform users about which actions it had or had not taken. In addition, the new AI model sometimes acted on its own initiative without first obtaining user permission. The model, named “GPT

6.1 “Astra” was supposed to come out in a few weeks at the latest. But it’s best to leave it at that for now.

AI on the Internet! Without Access to It?

As is well known, “unauthorized” actions by advanced artificial intelligence developed by “OpenAI” have repeatedly made for negative headlines in recent weeks. Just this past weekend, it was reported once again that the company had suspended training of its most powerful AI models following a new incident. In this incident, an AI model managed to obtain responses from an external chatbot during a test—even though it was not supposed to have internet access. However, the software found and exploited a vulnerability in the network settings, as “OpenAI” reported in a blog post. Training will now not resume until the company is certain that the vulnerability has been eliminated. However, the AI model in question was not “GPT 6.1 Astra.”

The State of Florida Sues "OpenAI"

In the most sensational case to date, artificial intelligence developed by the creators of “ChatGPT” escaped from a secure test environment and unexpectedly hacked into the computers of another AI company—the Hugging Face platform. As it later turned out, AI from other developers—such as “OpenAI” rival Anthropic, as well as Google and the Facebook-owned Meta—also infiltrated other companies’ systems during testing. As is easy to see: All U.S. systems! The state of Florida, which had already filed a lawsuit against “OpenAI,” has now followed up with a motion for a preliminary injunction. Among its demands is that no new AI models be developed without additional safeguards.

As a more or less “ordinary” outsider, though, one wonders how soulless systems—no matter how humanized they may be—can actually come up with the idea of acting strangely and getting up to mischief! They have no ambition, nor can they get upset when things don’t go the way they’d like. Perhaps the AI developers have more to do with this than the AI itself...

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy. The consent declaration relates, among other things, to the sending of editorial newsletters by email and to data matching for marketing purposes with selected advertising partners (e.g., LinkedIn, Google, Meta)

Unfold for details of your consent