The White House is reportedly considering vetting frontier AI models before release in response to growing security risks. The Trump administration has been non-interventionist on AI, but Anthropic’s new system Mythos proved capable of finding security flaws in almost all websites, and the firm has held back its public release as a result: The EU said it was in contact with Anthropic to get the region’s banks tested for security vulnerabilities. Other models will soon be similarly capable, and their developers may not be so cautious. The White House plan, modeled after Britain’s approach, would have officials review new releases to ensure they meet safety standards, The New York Times reported.
BitterGravity on
Under what constitutional authority?
Flat_Sail_7985 on
This just ties into heavily how they (The administration) is going to start vetting routers and similar consumer goods before they get released to the public.
Idk man. We are so screwed when it comes to this stuff long term. This is NOT me saying both sides are bad, or similar, but I think people are foolish if they think a future administration (be it republican OR democrat) will ever turn this power away. We have seen that every time the federal government gains ground, it, regardless of ideology, fights tooth and nail to retain it.
SizeFront7649 on
“Who won the 2020 election?”
Etnies419 on
Party of small government btw
jogarz on
This feels like another avenue for blatant corruption to me.
Automatic_Pepper_157 on
I started noticing Gemini’s google responses have begun being too generous to republican positions on the January insurrection.
disturbing
datums on
*The government needs to regulate AI, these billionaires can’t be deciding our future for us!*
Beyond the Republican bias they might want in the models, who’s going to be on this board reviewing them? “Oh, out experts from xAI have determined this model isn’t safe because it competes with Grok.”
TheBatz_ on
My ear has been pierced, but I can still hear the voice of the Party!
I’m a really strong supporter of precautionary principle type AI regulation, but I’m fairly confident that’s not we’re gonna get here…
Ok_Barracuda_1161 on
AI safety regulation is vital but of course I trust this admin will completely botch it. My main guess is that this will turn into rampant rent-seeking where AI models (like Mythos) aren’t safe for public release and can only be given to established firms who can meet expensive regulatory compliance
probablymagic on
The idea the president would prohibit companies from selling products until it had decided, by executive order with criteria id its own choosing is absolutely crazy.
Like, how about we shelf Anthropic models, because that guy reportedly hates Trump, we let Elon self-certify because he’s are bro, and we wait to see hoe much money Greg Brockman spends on the midterms to make a call on OpenAI?
14 Comments
The White House is reportedly considering vetting frontier AI models before release in response to growing security risks. The Trump administration has been non-interventionist on AI, but Anthropic’s new system Mythos proved capable of finding security flaws in almost all websites, and the firm has held back its public release as a result: The EU said it was in contact with Anthropic to get the region’s banks tested for security vulnerabilities. Other models will soon be similarly capable, and their developers may not be so cautious. The White House plan, modeled after Britain’s approach, would have officials review new releases to ensure they meet safety standards, The New York Times reported.
Under what constitutional authority?
This just ties into heavily how they (The administration) is going to start vetting routers and similar consumer goods before they get released to the public.
Idk man. We are so screwed when it comes to this stuff long term. This is NOT me saying both sides are bad, or similar, but I think people are foolish if they think a future administration (be it republican OR democrat) will ever turn this power away. We have seen that every time the federal government gains ground, it, regardless of ideology, fights tooth and nail to retain it.
“Who won the 2020 election?”
Party of small government btw
This feels like another avenue for blatant corruption to me.
I started noticing Gemini’s google responses have begun being too generous to republican positions on the January insurrection.
disturbing
*The government needs to regulate AI, these billionaires can’t be deciding our future for us!*
*Okay, but Donald Trump is the government.*
🤯
https://preview.redd.it/vbqz49a0gbzg1.png?width=1200&format=png&auto=webp&s=2cb34445a4c2160cfffd49a88481138df1bb69de
Beyond the Republican bias they might want in the models, who’s going to be on this board reviewing them? “Oh, out experts from xAI have determined this model isn’t safe because it competes with Grok.”
My ear has been pierced, but I can still hear the voice of the Party!
https://preview.redd.it/jnfstoqygbzg1.jpeg?width=1536&format=pjpg&auto=webp&s=6262c6eed1baca67c1b48836598d12675b87fd49
I’m a really strong supporter of precautionary principle type AI regulation, but I’m fairly confident that’s not we’re gonna get here…
AI safety regulation is vital but of course I trust this admin will completely botch it. My main guess is that this will turn into rampant rent-seeking where AI models (like Mythos) aren’t safe for public release and can only be given to established firms who can meet expensive regulatory compliance
The idea the president would prohibit companies from selling products until it had decided, by executive order with criteria id its own choosing is absolutely crazy.
Like, how about we shelf Anthropic models, because that guy reportedly hates Trump, we let Elon self-certify because he’s are bro, and we wait to see hoe much money Greg Brockman spends on the midterms to make a call on OpenAI?
Crony capitalism, full speed ahead!