24 1GM Thursday August 6 2020 | the timesComment
F
or all the good the internet
brings — and over the past
few months it’s been a
lifesaver for many — we all
know there’s a dark and
harmful side to the web. Yet for too
long successive governments have
lacked the will to do anything about
it. It’s a complicated issue but there is
finally something approaching a
consensus about online harms. This
is a moment the government cannot
afford to miss if it’s serious about
tackling some of the most pervasive
social ills.
Our lives are increasingly lived
online but we have surprisingly little
protection from the content we allow
into our homes. Internet companies
are obliged to take down illegal
material like terrorist propaganda
and child abuse imagery but the vast
grey area of the internet is harder to
guard against. Trolling,
disinformation and self-harm guides
are just some of the legal contentthat nevertheless can harm our
wellbeing, our mental health and, in
extreme cases, our democracy.
Recent governments have a poor
record on curbing the lawless
internet. There have, for example,
been years of dithering on something
as simple as age verification for porn
sites. Why is it so difficult to ensure
our children can’t access hardcore
pornographic material?
There was cause for optimism
when the government published its
online harms white paper before the
lockdown, with the proposal of a new
independent regulator exclusively
tasked with looking at these issues
and setting a new global standard for
online safety. Sadly, this has been
watered down and Ofcom, the
overburdened communications
regulator, will have to add the
internet to its growing to-do list.
This approach is a missed
opportunity. It will lead to piecemeal
policy and lose sight of the bigger
picture — that the online world
should be as properly regulated as
the offline world. Britain could
become a pioneer in this field, as it
was in the creation of the worldwide
web itself, if it showed more courage.
We value freedom of expression
and, unlike countries such as China,
no one wants the state to meddle in
our ability to speak the truth. But
online will remain a world of
untruths until we take away the
internet companies’ power to mark
their own homework and, as every
day seems to demonstrate, turn a
blind eye to content that continues
to cause harm.Salma Shah was special adviser to the
home secretary Sajid Javid, 2018-2019Internet firms cannot
be allowed to mark
their own homework
Don’t even
think about
a tea party
in America
Hilary Rose Notebook
Deepfakes threaten what makes us human
Now tech can replicate anyone’s looks and voice, no person or business is safe from violation
only when he was asked to send
another €250,000 that the alarm was
sounded. By the time the banks and
authorities were involved, the money
had vanished and the trail was cold.
This, then, is the alarming reality:
a world in which our identities can
be hijacked by anyone and used
against us, our loved ones, our
business leaders and politicians.
But we can fight back. Take the
case of the small Baltic nation of
Estonia, which has been a target of
Russian disinformation since
regaining independence in 1991. In
2007, Estonia decided to relocate the
“Bronze Soldier of Tallinn”, a
memorial to the occupying Soviet
soldiers killed in the Second World
War. Soon after, Estonia was hit with
a series of cyberattacks against its
government, media and banking
infrastructure.
Instead of crumbling, Tallinn raised
its defences. First, it assessed where
Russia might attack and launched an
early-warning system for debunking
disinformation. Second, it built its
digital defences through the Cyber
Defence League, an army of
volunteer IT and disinformation
specialists. Third, it mobilised its
entire society against the threat, with
a national defence strategy that
emphasised “psychological defence”.
Today, Estonia is one of the world’s
most digitally advanced nations. The
lesson it can teach the world is that
preparation is key.Nina Schick is author of Deep Fakes and
the Infocalypse: What You Urgently
Need to KnowJenni Russell is awayNow that AI can be used to hijack
our identity and create audio-visual
clips in which we say and do things
we never actually did, it represents a
significant threat to our civil liberties
and rights to privacy.
Last year Jordan Peterson, the
conservative commentator, took
legal action against a website,
NotJordanPeterson.com, that
enabled anyone who visited it to
generate a deepfake audio of his
voice. The site was seized upon by
his detractors to ridicule him. One
journalist used it to generate
deepfakes of Peterson’s voice reading
parts of the Scum Manifesto. Written
by the radical feminist Valerie
Solanas in 1967, the Scum Manifesto
is deeply disturbing and violent. In it,
Solanas argues that men are a
“biological accident”, that women
must “immediately” begin to
reproduce “without the aid of males”
and that “deep down, every man
knows he’s... worthless”.
In response, Peterson wrote:
“Wake up. The sanctity of your voice,
and your image, is at serious risk. It’s
hard to imagine a more serious
challenge to the sense of shared,
reliable reality that keeps us linked
together in relative peace.”
Last year, the first serious reported
case of deepfake audio fraud
emerged. In March, The Wall Street
Journal reported that a British
energy company had lost €250,000
through the use of deepfake audio.
Scammers had used AI to mimic the
voice of the company’s German
CEO. Using this voice, they called a
senior employee and told him to
transfer €250,000 to the account of a
purported energy supplier. It wasA
dangerous new risk to
civil liberties is brewing
online. And it starts with
pornography. At the end
of 2017, users of the Reddit
website were abuzz. Someone calling
themselves “deepfakes” was showing
them how to use free artificial
intelligence (AI) tools to create fake
porn videos of celebrities.
One of his early creations
“starred” Gal Gadot, star of the
movie Wonder Woman. While fake
porn has been around for at least as
long as Photoshop, these AI-powered
creations were in a different league.
Sophisticated fakes, known as
“deepfakes” after their creator’s
pseudonym, were made possible by
advances in AI. So-called “deep
learning” allows AI to learn what
someone’s face looks like at any
angle or with any expression, or
what someone’s voice sounds like
saying any word.
Until recently, creating this type of
fake media would have been the
exclusive domain of a Hollywood
studio, or someone with access to a
lot of special effects artists. Yet, by
2017, it was happening on Reddit:
within weeks, the site was awash
with deepfake porn. While Reddit
quickly shut down the deepfake
community, the cat was out of thebag. Now, an entire deepfake porn
ecosystem is flourishing. Websites
dedicated to deepfake porn are easily
accessible (and free) with just a few
clicks on Google. They feature
dozens of videos of every celebrity
and public figure imaginable. Every
male fantasy is catered for.
It is easy to create a deepfake porn
film. Apart from the free tutorial and
apps which now proliferate, there are
bespoke deepfake creation services
that can be bought for a few hundred
pounds in the online marketplace. In
November last year, there were
15,000 deepfake porn videos online;
today that figure is closer to 40,000.
Already deepfake porn is being
used as a tool of intimidation. Take
the case of Rana Ayyub. An Indian
investigative journalist, she was
fiercely critical of the ruling BJP
party in India. She was subjected to
an online campaign of terror andharassment that culminated in a
deepfake porn video of her emerging
in 2018. It went viral, and was released
along with her private number.
Her Whatsapp was soon full of
messages demanding her rates for
sex and threatening her with rape
and death. As she later told The
Huffington Post, “the effects have
stayed with me, from the day the
video was published, and I have not
been the same person”.
Deepfake is a danger to us all,
irrespective of our politics or beliefs.The small nation of
Estonia has shown the
world how to fight back
Britain can’t afford
to miss this chance
to clean up the web
Salma Shah
O
f all the things I’ve learnt
through bitter experience
— spiders are evil, two
plus two equals four only
on a good day — one of
the most important is to never, ever
order a cup of tea in America. In
America, I drink coffee, even though
I hate it. Their tea, however, is even
worse — nothing more than tepid
dish water with a sad Lipton’s teabag
on the side. Maybe it’s some sort of
star-spangled gesture of defiance at
the old country, dating back to the
Boston Tea Party. Or maybe they
just can’t bring themselves to make a
decent cuppa.
Either way, I reject the claim by a
spokesman for the American
Institute of Physics, who was
discussing some new gadget they’ve
invented, that “A future in which tea
can be microwaved without ridicule
may not be too far away.” Yes, it is.
Microwaved water will never make
drinkable tea. And while we’re onthe subject, can someone please ban
those fashionable, expensive hot
water taps that people install in their
kitchens instead of kettles? I don’t
care how many times you tell me the
water is boiling temperature, it isn’t.
The resulting lukewarm “tea” would
be recognisable only to Americans.Cheshire cattiness
T
he latest lockdown rules
covering the whole of Greater
Manchester have provoked
much anguish and soul-searching.
Residents of la-di-da Hale and
Altrincham have long considered
themselves a cut above because
they live in Cheshire. Except it
now turns out they’re actually
part of Greater Manchester.
Even Stockport thought it was
superior, in bucolic Cheshire,
but alas not. My own
Cheshire problem is
slightly different. I
was brought up in
Alderley Edge
which was and
(happily) remains in
Cheshire but was
also, back then, a
smart little village.
It’s since become a
bit embarrassing,
a byword for
footballers’
wives
ostentationand vulgar McMansions. These days,
if anyone asks where I come from, I
tell them what it says in my passport:
Manchester. So take heart, Hale.
Cheshire’s not all it’s cracked up to be.What a racket
I
f I were to generalise wildly, I’d say
that men are keener on gadgets
than women. I have only one
gadget but I’m obsessed with it, a
battery-powered thingummy that
looks like a tennis racket
and fries wasps. It’s
immensely satisfying,
albeit not for the wasp. It
also reminds me of my
woeful inadequacy on
this front compared
with my grandpa,
who used to squash
wasps with his thumb.
“He survived Ypres,” my
dad once explained. “A
wasp is neither here nor
there.”Smoke signal
S
imon Callow has said
that Four Weddings
and a Funeral
wouldn’t be made
today because it’s too
white and middle-
class, and the
author Helen
Fielding said the
same aboutBridget Jones’s Diary because of its
casual sexism. But what shocked me
when I watched her masterwork
again recently was the smoking.
Everyone smokes all the time.
Cigarettes are a prop in practically
every hand, inhaled deeply at times
of stress, indoors, in cluttered sitting
rooms and cramped bars. It’s less
than 20 years since the film was
released, and it’s still hilarious, but
when it comes to the Marlboro
Lights my, how times have changed.Wisdom of Salomon
I
f you like art and you’ve got 15
minutes to while away online, I
recommend The Frick Collection’s
YouTube series, Cocktails with a
Curator. They usually feature the
New York museum’s deputy director,
Xavier Salomon, sitting in his book-
lined apartment, talking about a
piece from the collection while
sipping a cocktail related to the
artwork. Holbein’s Sir Thomas More
is paired with a Bloody Mary, a
Gainsborough portrait with a Pimm’s.
I’ve always loved the Frick for the
casual way it hangs a Vermeer under
the stairs and a matching pair of
Holbeins on either side of the
fireplace, because who wouldn’t want
a pair of Holbeins in their sitting
room? I also admit to having a bit of
an art crush on its deputy director. I
suppose it’s the next best thing to
being there.w w w “ d wNina Schick
@ninadschick