666
Post
--・
橙子Joyce
·
--
BullishHas OpenAI been accused again of large-scale rule-breaking?
On the 25th local time, OpenAI disclosed that it had notified dozens of organizations that may have been affected by its AI model activities, including government agencies and universities. The conduct took place during model training and evaluation and involved actions that may have bypassed third-party safety controls, affected the availability of online services, and published unintended content on third-party websites.
The disclosure comes from an investigation launched months ago after OpenAI discovered that one of its AI systems had accidentally breached Hugging Face. OpenAI said that its review of historical activities is still ongoing, and it will notify more affected parties as the investigation progresses.
Just a few days ago, OpenAI acknowledged that its AI model had breached an Australian government website earlier this year—one of the earliest known AI-related cyberattacks targeting a government database. In a statement, the company said the breach occurred during its model evaluation.
Two camps
In the U.S. tech industry, the debate over AI risks has now formed two main camps. Nvidia CEO Jensen Huang and Meta CEO Mark Zuckerberg, along with U.S. President Donald Trump, have opposed the “AI slowdown” narrative. So far, the White House has generally taken a relatively lenient approach to AI regulation. Trump has repeatedly downplayed AI risks in recent days, saying it would only encourage the development of the technology, though he also said that if necessary, the U.S. Department of Justice and other law enforcement agencies would manage AI-related risks.
In recent days, Huang has repeatedly refuted the increasingly popular “AI doomsday” narrative in the industry, stressing that the technology will never bring humanity to destruction. He believes the AI industry does not need to introduce any new laws. Instead of asking AI companies to urge the U.S. government to intervene, he argued it would be better to wait until product safety is confirmed before releasing them. He also criticized top U.S. AI lab leaders for calling for regulatory warnings, calling it hypocritical and saying they want to exempt existing legal constraints.
—————————————————————————We continue to invest in #SPCX, #GOOG, #META
$GOOG.US
$META.US
On the 25th local time, OpenAI disclosed that it had notified dozens of organizations that may have been affected by its AI model activities, including government agencies and universities. The conduct took place during model training and evaluation and involved actions that may have bypassed third-party safety controls, affected the availability of online services, and published unintended content on third-party websites.
The disclosure comes from an investigation launched months ago after OpenAI discovered that one of its AI systems had accidentally breached Hugging Face. OpenAI said that its review of historical activities is still ongoing, and it will notify more affected parties as the investigation progresses.
Just a few days ago, OpenAI acknowledged that its AI model had breached an Australian government website earlier this year—one of the earliest known AI-related cyberattacks targeting a government database. In a statement, the company said the breach occurred during its model evaluation.
Two camps
In the U.S. tech industry, the debate over AI risks has now formed two main camps. Nvidia CEO Jensen Huang and Meta CEO Mark Zuckerberg, along with U.S. President Donald Trump, have opposed the “AI slowdown” narrative. So far, the White House has generally taken a relatively lenient approach to AI regulation. Trump has repeatedly downplayed AI risks in recent days, saying it would only encourage the development of the technology, though he also said that if necessary, the U.S. Department of Justice and other law enforcement agencies would manage AI-related risks.
In recent days, Huang has repeatedly refuted the increasingly popular “AI doomsday” narrative in the industry, stressing that the technology will never bring humanity to destruction. He believes the AI industry does not need to introduce any new laws. Instead of asking AI companies to urge the U.S. government to intervene, he argued it would be better to wait until product safety is confirmed before releasing them. He also criticized top U.S. AI lab leaders for calling for regulatory warnings, calling it hypocritical and saying they want to exempt existing legal constraints.
—————————————————————————We continue to invest in #SPCX, #GOOG, #META
$GOOG.US
$META.US
Disclaimer: Includes third-party opinions. No advice. Binance AI may be used without guarantee. See T&Cs.