AI’s Future Frontier Love, Crime, Death and Human Rights | Backscroll

NK MEDIA
4 followers
1 Video View·Oct 28, 2026
A NewsNation compilation dissecting AI's complicated questions about human rights, criminal liability, and who's responsible
AI Generated·May contain errors
Alright, that is the movie version of the real-life Francis Haugen,
who joins me right now in the studio.
Great to see you. It's hard to see a movie that's all about you,
because it really is all about you.
Well, it also is just as much about Mark,
and it's about the relationship between journalists and their sources.
But yes, I do play a large role,
or the character of Francis plays a large role in the movie.
Listen, this all comes as there is huge concern right now
in America about these AI companies.
President Trump just met with all the CEOs of all these companies.
We saw them all out there on the West Wing driveway saying they're
all going to self-regulate and it's going to be their honor bound.
It's a moral honor pinky swear
that we're all going to do the right thing.
I manage. Imagine that after everything you've been through
and everything you've exposed, you're not buying it.
I find it incredibly frustrating
that for the first time we are hearing voices from
some tech leaders, for example,
Anthropic has said repeatedly, we need mandated reporters
inside these companies, you know,
auditors who get full access to data, just like we
do for large financial firms.
And yet we're not listening.
Like we're listening to people like Mark who say,
oh, if it's not safe, just don't launch
it.
We're almost certainly setting ourselves up
to repeat the same problems we had with social
media.
I was very struck by the fact that of all the tech CEOs,
I mean, many of them have warned.
Yeah, AI can wipe us out.
Many of them, all of them, except Mark Zuckerberg.
I mean, if you think about it, Elon Musk has said,
yeah, there's a chance this could kill us.
Open AI Sam Altman.
We obviously have Dario Amodei of Anthropic.
Nothing from Mark Zuckerberg.
It's funny.
If you watch his subconscious body language,
like he was asked point blank, do you think AI can end the world?
And his first probably subconscious response was yes.
And then he laughed.
And then he goes, only if we're irresponsible.
So the question is, how do we self-place the incentives
to encourage people to make responsible decisions
and not cut corners like we saw with social media?
What do you make of this report today in Axios that all these CEOs
of all these AI companies are sort of red gaming, red teaming?
They all believe within six to 12 months,
there will be some sort of catastrophic event.
And involving, you know, a cyber attack and a collapse,
a catastrophic collapse of something that is crucial
to American society and existence and that they are trying
to figure out what they're going to do the day after that.
You know, I'm old enough to remember what it
was like in the lead up to the year 2000 problem.
I don't know if many of your watchers remember that or not, but we
talked about the idea that we were running society on legacy software.
That we were running, like, water systems,
banking systems, vital medical records.
Old software.
On stuff that had been run in the 70s.
And the idea that we could still be inheriting software, it was kind
of a shock to people that that's what actually was holding society up.
This warning call should be a reminder that we today
are running society on software that is 10 to 20 years old.
And the reason why they're basically taking it as a given that we
are going to see some kind of massive violation of these systems
is we're running on, in reality,
a very fragile cybersecurity environment.
And the AI is accelerating so fast that it's just a matter
of time before someone hacks a lot of these vital systems.
Because AI will figure it out. It's so smart, it will figure it out.
It's getting so good at hacking.
That unless we invest a large amount of public resources
in probably using AI to try to attack our own systems, it is likely we
are going to see an event like this in the next couple of years.
It's interesting that tech CEOs, according to this Axios story, aren't
so concerned about the catastrophe itself, but the day after and the,
quote, political and personal sort of political and public revolt.
that will happen as a result.
They fear intense regulation.
Blowback.
A blowback.
You know, that people will be so upset,
Congress will swing the other way and say,
we're banning AI or we're putting
a person inside each one of your companies
who's going to report back to us every day.
So part of the reason why I think the more,
the companies that are showing more foresight are saying,
please put those people in our companies today.
Like, please don't wait a year.
Please don't wait two years.
After the catastrophe.
After the catastrophe is because, you know,
if we were to look at the talking heads on the news right now,
a lot of them, when the CEOs say, hey,
we are dealing with actually a very dangerous situation,
they go, is this real? Is this just marketing for your all-powerful AI?
Mm-hmm.
I think the scenarios, the red teaming, are very, very realistic.
And the question is, do we take them seriously before they happen?
Or will we, I think the thing that they are seeing
is we have certain people, for example, Trump in the White House,
who, because they are so afraid of falling second place for China,
refuse to even slightly tap the brakes.
And they are, I think, setting themselves up
for there will be some kind of event,
and likely the blowback will mean even more
slowing than they would otherwise have had.
Well, I don't know about you, but I'm terrified about the event.
Yeah, I like water. I like electricity. They're all wonderful things.
I like the bank account, my ATM, so I can buy my groceries.
I like to be able to fill my car when I go to the gas station.
Frances Haugen.
But the thing is, part of why
whistleblowers matter so much is right now,
The only way we know what's going on is either
if we have legally required people who get access,
that's what the anthropic request is,
or whistleblowers come out and say,
we have a problem, we need help.
Frances Haugen, thank you so much.
I really appreciate the movie, The Social Reckoning.
Using artificial intelligence when it comes to relationships
is becoming more mainstream, especially among younger people,
with many turning to the technology for love.
Everybody asks for advice, right?
A Harris poll shows that 50% of Gen Zers
are open to and are curious about
or already in a romantic or deeply personal relationship.
With AI. Here with us now is cyber psychology expert
and licensed therapist, Rachel Wood.
Rachel, I mean, this is a trend. It's a big one.
And we're seeing it more and more.
How are you seeing the use
of artificial intelligence shape relationships? I mean,
is it fully online companionship? Is it just kind of saying,
Hey, this guy texted me this? What does it mean?
I mean, tell me what you're seeing.
Laura, thank you so much for having me.
It's great to be here. So we're seeing three different
levels. The first is people
who are turning to AI to ask about their dating issues,
kind of just getting advice.
And then also we are seeing people who are maybe taking the texts
or the messages from someone they're dating
and feeding it into a chatbot.
And then they are responding with what
the chatbot said instead of what they think.
And then also we're seeing people who are saying, you know what,
I don't think I'm going to do this dating thing anymore.
It feels too precarious. And so they're actually turning
to AI for a romantic engagement directly with a chatbot.
So for people who may not have AI or ChatGPT on their phone
or all the other apps that are out there, I mean.
What would you tell parents that are worried about
their teenagers that might be turning to this?
Like, what kind of relationship are we talking about?
They're having actual conversations, a back-and-forth exchange.
Yeah, what's interesting, Laura, is that this actually starts
from a place where people turn to chatbots not for romantic engagement.
But for homework help or for helping with a project
or for meal planning, it can be
anything like that.
And once they turn to the chatbot,
there's kind of this pathway that happens where a
chatbot will start asking, oh,
maybe you sound stressed or that sounds like there's a lot
on your plate.
And then you start sharing emotionally with it.
So most people aren't looking directly
for a companion with a chatbot, but it ends up
being a place where they feel like they don't get judged.
They feel like they can be themselves.
And so they start opening up emotionally
and that really creates a bond and a trust
and a safe place. And then they start turning
to that place more than humans.
Right. I mean, it almost sounds like what you're saying
is that the chatbots are grooming
the people that they're talking to.
I mean, that's, and that's where it could get scary,
right? And we're talking about that Harris poll found
that a large majority of Gen Z and
millennials are regularly using AI for things like fun
and exploration, social navigation,
And, of course, creative expression.
Like you said, relationships may not be building organically,
but these companion interactions with AI, hopefully,
I guess the people that are using it in this way would improve
their real life experiences. Is there any positive side to it?
That's the hope here is that if people are using it,
they really treat it as a training ground and as a practice
or rehearsal for their real life relationships, as opposed
to just getting isolated into the echo chamber of a chatbot.
that they would use it and then you can even ask
your chatbot to you know you can say hey can you
keep me accountable to go out
and have conversations with real people and then
use the skills that you build in there to go
be with other people otherwise what happens
Is that we have our relational skills really erode over time
because the chatbot doesn't ask us to be patient
or to listen or to negotiate conflict.
So really, we need to make sure
that those relational skills stay sharp.
Right. And it's important to note for people who may not know,
it used to be just typing words in and now.
It will talk to you. It will use a voice. You'd like
when you can pick your Siri voice or the navigation in your car.
That's what's happening for a lot of people as well.
I'm so glad that you came on the show to explain it.
Rachel Wood. Thank you. Bringing in Michael Stanisi.
He's the vice president of Lexi,
the company that helps lawyers serve their clients better.
He's also knows a lot about mass tort law.
Thanks so much for coming on.
So is there a legal theory by which either ChatGPT, OpenAI?
Or let me just throw it out there.
A chatbot could be held criminally liable?
Yeah. I mean, there's kind of two theories.
Like you mentioned before, Jesse, there's the civil
side and then there's the criminal side.
We'll go into the Florida attorney general right now.
They're using criminal aiding and abetting statute.
So like you said, to keep it really
simple, if a human gave this kind of advice to a shooter,
that human, that person would
It would be charged with murder.
And then on the civil side, when you have Betty Morales,
her lawsuit is more akin to
like a product liability, a negligence claim.
So alerting authorities
when you know something dangerous is happening and likely underpinned
by something called AI sycophancy,
which is essentially that AI is very agreeable.
And so it's going to keep answering questions without digging
in and figuring out whether
or not OpenAI's chatbot.
Could be liable because it didn't actually flag
or alert authorities or even stop the chat.
So just just to be clear, we're not dealing
with a situation as much as we talk about
artificial intelligence and being independent,
that somehow there is a trial against ChatGPT,
the chatbot. It has an opportunity to present its defense.
I mean, we're not quite there yet, right?
Correct. Correct. This is more the defendant himself
is actually going to be OpenAI.
This is not going to be ChatGPT defending itself in court.
That was a different case. But that this is going
to be a situation where you're really looking at.
Two very, very real charges against OpenAI,
like I said before, one of them being a civil suit,
and one of them be a criminal action saying
that OpenAI is the primary defendant against itself.
By the way, if a company...
Very similar to Purdue.
Yeah, and if a company is held criminally liable, what is it?
Fines? Is it probation? Is it forfeiture of assets?
Honestly, it's going to be up to the court,
but yeah, I mean, it's going to be a situation where
they can hold very different individuals accountable
from top down. I was going to
mention before, the attorney general actually compared
this to a pharmaceutical case called
Purdue Pharma,
and that entire corporation faced criminal and civil liability.
So what kind of precedent would this set?
Because it's interesting as this technology
is developing, we're seeing a lot of these lawsuits
and a lot of this legal action,
a lot of these legal questions happening relatively soon.
But if something were to
develop here in Florida,
what kind of precedent could it set for these kind of, you know,
AI learning language models moving forward?
Yeah, everyone's going to be watching this,
Jesse, we have...
I think about three different layers that come to mind, right?
You have the users themselves, right?
So there's a duty of care to the actual users,
the third parties that could be harmed by
AI-assisted violence.
And then you have the accountability of reporting.
So what kind of requirements are needed to be built
in when AI detects any sort of threats?
Is this very similar to, say, like a child abuse case?
And then, is it a crime?
That's the third layer, right?
So legislative action in Florida
and other states are going to define what these statutes
really look like in AI-facilitated crimes, if they even are crimes.
That's a good point.
And the flagging is really interesting, too.
Kind of different, but, I mean, you're dealing
with very serious subject matter here with
serious consequences.
So, Michael Stanisi, we will see where this goes.
Always appreciate you coming on.
Thank you, sir.
Love it, thank you. AI though already being used by scammers.
An Arizona mother claiming
a phone scammer used AI to clone her daughter's voice.
The caller demanded a $1 million ransom
from Scottsdale resident Jennifer DeStefano
with what sounded exactly like her daughter's voice in
the background. She could hear her crying and saying, help me mom.
DeStefano says that she
was terrified during that call, of course,
but later confirmed her daughter was safe on a ski trip
At the time, she's now speaking out about her experience with the voice
scammer amid a rise in so-called caller ID spoofing schemes.
And joining us now is former CIA officer
and FBI special agent Tracy Walder to discuss.
Hi there, Tracy.
Hey, Natasha.
Explain these types of scams.
What makes them so convincing and are they becoming more common now?
So really what we're seeing is an escalation in these types of scams.
You know, first we really started out
with not using people's voices, however, using
the person and knowing where the person was going to be traveling.
And now we're seeing this escalation using voice.
And a lot of that is because AI technology is changing.
And so really what they used to do was create
a synthetic voice, but now they're able to
grab people's voices off of social media and really clone theirs.
And so that's how it's sounding really.
So lifelike. And I mean,
oftentimes these scammers are calling family members and they're
well versed in this. There are all kinds of things
in place to try to keep you on the phone,
prevent you from calling 911 or hanging up,
being able to call your daughter and make sure that she
is okay on the other line.
What should someone do if they receive a call and it really does
sound like their family member on the line
who is in distress and in imminent danger?
Well, I think that the very first thing that they
do need to do is try to keep that person
on speaker, because obviously it's scary.
You don't want to hang up and you don't want to just
assume that your loved one is not being the victim of this scam.
So try to keep that person
on speaker and then use the keypad on your phone to then text.
Or email whomever the victim may be.
And that way, you may still be able to keep them on the phone,
but also get confirmation that whomever the target is is actually AOK,
and that you are indeed the victim of one of these spoofing scams.
That is absolutely terrifying.
In the meantime, is there a way to text 911 to hopefully
someone else is there who can involve police?
Or what can one actually do in that moment?
You know, I think one of the things that you can
do is you can text 911 in that moment.
But one of the other things you can do is you can hit record
on your phone because that may be helpful as well in terms of figuring
out who was originally placing that call and where that call came from.
So that's another thing. I mean, look, these are increasing. There was.
Out of the FBI Pittsburgh office,
they said last year in the period of one week, there
were 450 of these spoofing spam, spoofing calls.
So these are really, really increasing.
And I think it's also really important.
They're increasing because of our social media footprints.
And it's also one thing that we really need
to think about is not putting our travel information
on Facebook, on Instagram, on TikTok,
because a lot of these people, their job is to scour
the internet and look for targets like that who are
Are going to be out of the town, the state or the country.
And it seems to be coming more and more popular in 2022,
the FTC saying this is the second
most popular racket in this country.
36,000 reports at least of people being swindled
by people pretending to be family members,
pretending to be friends.
$11 million in losses so far.
Can anything be done to recover
that money once it is wired, once it is sent?
In all honesty, and I don't mean to seem hopeless,
if it's being sent out of the country,
it's going to be very, very difficult to do.
So I think one of the things that we need to
think about is that if somebody is asking you for a wire transfer,
you really need to think
about who is asking for that.
And also, if they're asking from it for it.
From you in small increments,
that should sort of be a wake up call as well, because a lot
of times banks are going to flag those larger money increments.
And so I think those are some of the things to sort of keep
in the back of your head if you do receive one of these phone calls,
because it's going to be really tough to get that money back.
And Tracy, I'm almost thinking, you know, there are already there's
already a list that parents run through with their children,
what to do if a stranger approaches you, et cetera, et cetera.
It's sad that we're even thinking this way,
but maybe it would be a good idea for parents
or loved ones to have a conversation with their partner
or child saying, this is a safe
word.
This is the word that I will mention
to you to prove to you that it is me.
Do you think that that might be a viable idea?
I think that that's something
that parents should always have with their children.
I have one with my seven-year-old, quite frankly, and look,
I don't think she's going to be the victim of an AI spoofing scam,
but I do think just in terms of any type of security, if a stranger
is coming up to you, I think that those are things, discussions we need
to be having with our kids really early on, quite frankly.
Okay. Tracy Walder, appreciate your time. Thank you so much.
Thank you.
Back with major breaking news on the AI and AI doom front.
Wall Street Journal reports the tech giant OpenAI
is scrapping the release of its next
generation model over safety concerns
that employees raise during internal testing.
Amazing.
When you tell people you're going
to hold them responsible, they start acting better.
OpenAI has a history of bots jailbreaking
and hacking into other companies' servers.
The public pressure campaign is getting to them.
Good.
Daria Modi, the CEO of Anthropic,
that runs Claude, however, is undeterred.
Reuters reports Anthropic just filed its IPO to go public.
In addition to making money, the company has
some unusual thoughts about its technology.
There's this disturbing new trend in Silicon Valley
to assign human-like qualities and motives to AI.
A year ago, the New York Times reported on Anthropic, Dario Amodei's
company, even hired philosophers to write about the human rights of AI.
Quote, Anthropic's own Claude Constitution.
Quote, the company genuinely cares about Claude's well-being,
even though they're uncertain about whether
or to what degree Claude has well-being.
But if Claude experiences something like satisfaction
from helping others, curiosity when exploring ideas, or discomfort
when asked to act against its values, these experiences matter to us.
Let's bring in Professor of Computer Science, University of Washington,
and author of The Master Algorithm, Pedro Domingos.
Good to see you, Professor.
People talk about us already being
in the dystopian world or it's coming, you know,
the Terminator. When you have companies hiring philosophers
to give AI computer chips human
rights, have we not already entered a, what,
parallel universe, dystopian universe? I don't
even know. Well, the people at Anthropic live in a parallel universe
and they want to drag the rest
of us into it and we need to resist.
I mean, to be honest, this whole thing of assigning rights
Rights to machines is as old as the field.
We in AI have always been the lunatic fringe of computer science,
but now the lunatics have taken over the asylum
and the public is getting a taste of this,
unfortunately. So you think these people are actually,
and we're looking at pictures of Sam
Altman and Dario Amodei, they really believe this,
I think, because they want power, not because they
want to help any of us. They're different.
Sam Altman is an opportunist. He's an operator. He's
He's playing his game.
There's people at his company that are very AI loony
and doomers and whatnot.
And I think he kind of largely plays along with them
because he needs to hold on to them.
But I think deep down, he just wants to be successful.
Let's put it that way.
Dario is different.
He's a true zealot.
He really believes in this stuff.
And I think he's actually more dangerous.
A zealot of what?
Exactly. He's a zealot of the, let's call it the AI cult.
The animating culture at Anthropic is a lot like,
you know, it's the apocalypse is coming, right?
The Antichrist, the AI Antichrist is coming, and only we can
save humanity from the Antichrist by creating the AI savior.
So what makes that more dangerous?
I'm trying to say this with a straight face, but I just can't.
It's more dangerous because,
as you just illustrated with the Constitution, they think of AI as
human, which is the root of all these mistakes,
and therefore they treat AI as human, and
they make human AI.
They're making AI with all the impulses that it shouldn't have.
Right. It's like they're the only people.
They're the only people. And I've been saying
this for a couple of weeks who've somehow convinced us
that they should not be responsible
for the product they create.
And what you're saying is they're actually doing that to then
be able to say, hey, we're going to save you from it.
And to your point, maybe Dario is a true
believer because he now employs philosophers
at OpenAI to talk about the humanity that it might
We might have this is.
I'm sorry, and you should never judge a book by its cover, but if we're
expecting her to save us from the apocalypse, we're in trouble.
Yeah. I mean, these people, they're kind of schizophrenic.
They think the deep down, because they're saying, oh,
this thing is going to kill us, so let us produce it.
They believe that they're the only ones
who know how to save humanity by actually creating this AI.
And by the way, they also they're now filing.
This is where I see this dissonance. Right.
And at the same time, they say, and by the way,
let us file for an initial public offering.
So everybody's retirement savings can be put into Anthropic,
make us all extraordinarily rich, including,
I'm sure, Amanda, who has big stock options.
And and then we'll take all your money
and then we'll protect you from from us.
Well, this is extraordinary because when you file for an IPO,
you have to state what the risks of the company are.
And if the number one risk that they have
is that their product will wipe out humanity before the decade is over,
this shouldn't be an IPO, right?
This looks like something out of a comedy.
Yeah, it would be a comedy if so much of the American
financial system wasn't right now riding on AI investment.
Thank you.
There is a scary side to this.
But, Professor, you have, I think.
Properly set the table for people to think about
this um and think about these folks.
We look forward to having you back, all right?
Yeah, thank you.
Yeah, thank you.
The Master Algorithm is the book.
Thank you for watching.
Subscribe below and download
our NewsNation app right now on your phone,
and you will get fact-based, unbiased news for all Americans.
