Microsoft co-founder and billionaire philanthropist Bill Gates on Wednesday called on the U.S. to be a leader in the global regulation of artificial intelligence growth but warned that getting countries to agree on a global framework for AI regulation could be “more difficult” than Cold War-era negotiations on limiting the development of nuclear weapons.
uncertain
countries → call → weapons
“This is more difficult than that was,” Gates told NBC News’ “Meet the Press,” during a wide-ranging discussion about AI that aired on Sunday.
asserted
that → tell → Sunday
His remarks came as he spoke about President Donald Trump’s stance against regulating AI.
asserted
he → come → AI
The president earlier this month rejected calls for the federal government to impose guardrails on companies building AI models, saying in a post on Truth Social, “WHOEVER WINS AI, WINS!
asserted
WHOEVER → reject → AI
We are leading China, and all others, and will continue to do so.”
asserted
We → lead → China
Gates, in Wednesday’s interview, argued that regulating AI companies by imposing monitoring systems would not hamper the U.S. in any race with China to develop superior technology.
asserted
regulating → argue → technology
“What he said is fine, but I think after he hears from me and other people, including that this does not handicap us in whatever he thinks the nation-state race is, that, you know, we will come to a consensus,” the tech mogul said.
asserted
mogul → say → consensus
Gates later pointed to other times the U.S. has led the world as an impartial figure in coming up with broad agreements to protect humanity.
asserted
U.S. → point → humanity
“Since World War II, the U.S. should be very proud that when there have been global challenges, we, in a way that was benign, you know, where we had friends and allies coming along with us, you know, aviation safety, ozone layer, nuclear weapons, a lot of things where we were played the fair actor, not aggrandizing whatever our unique strengths were,” he told moderator Kristen Welker.
asserted
he → know → Welker
“And AI is one we’re going to have to go back to this, ‘Hey, humanity is all in this together.’”
asserted
humanity → go → this
The Microsoft co-founder’s comments come as concerns about AI reached a fever pitch after Jacob Coxon, a whistleblower who’d recently resigned from Anthropic, made public comments saying, “The people building AI earnestly believe that it could kill us all by the end of the decade.”
uncertain
it → come → decade
Days after those public remarks, two of the nation’s leading AI company CEOs — Dario Amodei of Anthropic and Sam Altman of OpenAI — agreed that there is a need to slow the pace of AI development amid fears that AI agents will soon be capable of “superintelligence” that outpaces the bounds of human thinking.
asserted
that → lead → thinking
In a blog post, Amodei wrote that frontier AI companies must slow down the pace of improvements to AI models but only to a point where the companies still lead ahead of China.
asserted
companies → write → China
He also urged global coordination between democratic and authoritarian governments to create regulations.
asserted
He → urge → regulations
In recent weeks, both companies have announced plans to track the behavior of AI agents after OpenAI flagged multiple incidents in which their agents behaved in a concerning way, including by hacking an outside company without being told to do so.
asserted
agents → announce → company
In a post on Truth Social ahead of a meeting with Chinese President Xi Jinping on Thursday, Trump wrote, “A big day with President Xi of China.
asserted
Trump → write → China
Super Intelligence (SI) will be a big topic of discussion, but I want to leave it exactly where it is.
asserted
it → want → it
Addressing immediate concerns about AI
While many have been calling for federal or global regulations that would limit AI from destroying humanity, Gates in the “Meet the Press” interview said he believes there should be guardrails instituted now to prevent mass casualty events that could be coordinated with the help of AI as soon as today.
uncertain
that → address → today
“The most dangerous thing we’re facing right now is people with bad intent using AI.
asserted
we → face → AI
You know, so somebody who wants to defraud an individual or wants to shut down the electric network or, you know, scramble bank accounts, they are now empowered,” the philanthropist said.
asserted
philanthropist → know → accounts
He added, “historically, countries, you know, had cyberattack teams and they could think of things like bioterror weapons.
uncertain
they → add → weapons
Now, small groups with AI can do what only the biggest countries could do.
uncertain
countries → do → what
So it’s really the bad intent plus AI that looms as a major threat.
asserted
that → ’ → threat
Unless governments insist on various safeguards, you know, we will see a lot of big, negative things because we’re not safeguarding.”
asserted
we → insist → things
Gates specifically said monitoring, “where you’re watching what the AI is doing,” is an important safeguard to introduce in each AI lab.
asserted
AI → say → lab
“Law enforcement is going to get called up, you know, on all these fraud events or, you know, transmission shutdown events.
asserted
you → go → events
And their ability to see what’s going on is going to be nil, unless the monitoring and safeguards are built into the AI,” he added.
asserted
he → see → AI
Plus, Gates said, these kinds of safety features can be added into current AI systems without slowing down the pace of AI development.
asserted
kinds → say → development
“That’s another debate — of should we slow down or not,” he said.
asserted
he → ’ → ?
“But deciding to put the safeguards and monitoring in, that’s a choice we have.
asserted
we → decide → safeguards
And the only question is do we do that before or after there’s some large-scale cyber attacks or bioattacks.”
asserted
we → do → that
One bipartisan proposed piece of legislation in Congress is one that would impose requirements for a “kill switch,” or a way for AI companies to immediately suspend or shutdown their AI services in the event of a dangerous situation.
asserted
companies → propose → situation
Gates told NBC News that, “it’s not enough to have a kill switch,” and renewed his calls for internal monitoring and safeguards for AI agents to be imposed.
asserted
monitoring → tell → agents
“I would never want to say I’m against a kill switch,” he added.
asserted
he → want → switch
“But when you’re trying to moderate the bad behavior, you need insight and records of what’s being done.
asserted
what → try → insight
So the kill switch alone would not prevent these tragedies.”
asserted
switch → prevent → tragedies
The Gates Foundation and AI
The Microsoft co-founder, who now leads the Gates Foundation, also spoke about the group’s work expanding AI access to countries where English isn’t the primary language.
asserted
English → lead → countries
“AI models have gotten very good at recognizing English and about a dozen other mainstream languages.
asserted
models → get → English
And so as you talk about your health, it’s going to accurately record that and be able to help you.
asserted
it → talk → you
In African languages, the error rate will be about 10 times higher,” he said, adding that the error rate can be “very dangerous.”
asserted
rate → say → languages
…and 17 more, not listed.
The Aporia
