The Facebook Papers reveal the limits of
regulation. It’s time to think bigger
Lawmakers across the Western world are thinking too
narrowly in how they police social media companies.
The documents show that lawmakers are still not
thinking big enough. | Josh Edelson/AFP via Getty Images
BY MARK
SCOTT
October 26,
2021 7:23 pm
If the
revelations from the so-called Facebook Papers showed us anything, it's that
new rules for policing what's posted online are needed. And fast.
The bad
news: That's not going to happen — at least not in ways that will make a real
difference.
Even before
the inside look into Facebook's handling of online content, first reported by
the Wall Street Journal and then by multiple media outlets, including POLITICO,
policymakers in the European Union, United States and elsewhere were putting
together proposals to force global platforms to be more accountable for what
people posted online content, including potentially hefty fines when things
inevitably go wrong.
But as
lawmakers battle it out over the EU's Digital Services Act, the United
Kingdom's Online Safety Bill and the U.S.' stumbling efforts to revamp its own
law, known as Section 230 of the U.S. Communications Decency Act, none of this
legislation will tackle the underlying issues laid bare by what has been
revealed about Facebook's oversight of online content.
What the
internal research, emails and insights show is a complex web of interconnected
levers, which — taken together — have created arguably one of the most profitable
businesses the world has ever seen, and built an online behemoth that shapes
how almost all of us live.
As I waded
through these internal documents, what struck me was how all-encompassing
Facebook is. Engineers routinely tweaked content algorithms to prioritize some
content over others. California-based executives regularly made choices that
affected people in far-flung locations in almost real time. What the tech giant
knows about people's daily digital habits — based on minute-by-minute monitoring
of everyone's online interactions — would make George Orwell's "1984"
look like a children's bedtime story.
The
documents show that lawmakers are still not thinking big enough.
So far, the
legislative proposals focus too narrowly on taking small bites at a much bigger
problem. They would force social media companies to open up their data to
outside researchers; limit how some political ads can target would-be
supporters online; and demand companies make public their plans for tackling
existential risks like when elected political leaders peddle misinformation and
hate speech on their platforms.
But it's
not enough to legislate on parts of what social media companies do. Until
lawmakers take a step back and think more broadly about the rules overseeing
online content — and its links to other digital priorities like privacy and
competition — the abuses outlined from the internal documents are unlikely to
go away.
For its
part, Facebook has defended its oversight of online content, saying it has
spent billions of euros, hired tens of thousands of content moderators and
changed its algorithms to tamper down hate speech, misinformation and divisive
political material and protect its 2.4 billion users worldwide.
What
content to police?
So where
have things gone wrong? Let's start with arguably the most far-reaching online
content rules — the EU's Digital Services Act.
These
proposals tick a lot of boxes.
Under plans
expected to become law sometime next year, Facebook and others will be required
to conduct regular risk assessments on potential problematic hotspots on their
platforms, and hire independent auditors to make sure they're not cheating.
Some outside groups will have (limited) access to internal social media data to
analyze what's going on, while the bloc's regulators will be able to fine firms
up to 6 percent of their annual revenue, amounting to billions of euros, if
they don't comply with the rules. So far, so good.
But where
this legislation falls is that it narrowly focuses on illegal content. Other
material, including suspect social media posts that nudge up to the line of
illegality, but never cross it, remain out of scope.
What
internal Facebook documents show is that such “harmful non-violating
narratives," to borrow from the company's own language, have repeatedly
played a significant role in fomenting division and, in some cases, offline
violence.
In the days
around the January 6 Capitol Hill riots, for instance, the tech giant's
engineers didn't take action against posts that questioned the legitimacy of
last year's U.S. presidential election, even as such material was being used to
promote political attacks in Washington. Another example: content that verged
on anti-vaccine misinformation, now banned on the platform, was also not
subject to review, despite that content fomenting online conspiracy theories
around COVID-19.
European
lawmakers said they had to draw the line somewhere, and expanding the scope of
the bloc's proposed content rules to include harmful, but not illegal, material
would have proved unwieldy. But this gap in the legislation — one that will
allow reams of problematic content to remain online and off-limits to
regulators — is a major blind spot in Europe's push to police social media
Political
ads, politicians and the media
In London,
the country's lawmakers want to solve that problem by regulating harmful online
content, even if it is legal.
Under the
U.K.'s Online Safety Bill, which is expected to voted on before the end of the
year, social media companies will have a so-called duty of care to protect
users from such problematic content — even if it doesn't break the country's
existing hate speech laws. Fines totaling 10 percent of annual revenue could be
levied for noncompliance, while social media companies' executives could even
be sent to prison if the worst-offending material is ignored.
But, again,
London's plans are flawed too.
Negotiations
are still underway. But the proposals currently exempt online political ads, as
well as posts from politicians and publishers, from any form of content
moderation.
Facebook's
internal documents revealed that paid-for partisan messages played a
significant role in promoting divisive, nonpaid content across the platform.
Elected officials and media organizations (some of which have been created by
political groups to push their own agenda) also had a hand in fomenting
distrust — and the problem went well beyond former U.S. President Donald Trump.
Frances
Haugen, the Facebook whistleblower, said these limits could undermine the
U.K.'s upcoming content rules by failing to confront what was really going on
within Facebook.
"I am
extremely concerned about paid-for advertising being excluded because
engagement-based ranking impacts ads as much as it impacts organic
content," she told U.K. lawmakers on October 25. "It is cheaper,
substantially, to run an angry hateful divisive ad than it is to run a
compassionate, empathetic ad."
Et tu,
Washington?
Across the
Atlantic, content moderation legislation continues to move at a glacial pace,
even as Haugen meets with U.S. lawmakers also wading through Facebook's
internal documents looking for answers.
New rules
aren't likely anytime soon. But even the proposed bills — most of which would
make social media companies more liable for what's posted online and force
greater transparency over how decisions are made in Silicon Valley — fail to
tackle how these firms' algorithms often push harmful, viral content over more
staid material. Other bills specifically looking at monitoring such algorithms
face little, if any, chance of becoming law.
No plans to
regulate, or even monitor, such systems are expected to pass in Washington for
the foreseeable future.
Internal
Facebook documents, from 2018 and 2019, respectively, highlighted how
Facebook's automated systems prioritized negative content because it was more
likely to go viral — and therefore keep people on the social network. No likely
law in the U.S. would deal with this underlying issue.
Until
Western policymakers start thinking bigger, these fundamental flaws with
existing online content proposals will not be fixed — and still leave people
mostly unprotected from the dangers that the Facebook Papers have made
painfully clear.
Mark Scott
is chief technology correspondent at POLITICO.
Sem comentários:
Enviar um comentário