I live in a part of the world where creating content is king no matter what your sexuality may be.
It is becoming an increasingly familiar story for LGBTQ+ creators. Years of work disappear overnight. Entire YouTube channels vanish without warning. Videos are removed. Monetization is stripped away. Appeals are rejected or met with silence. Then comes the same explanation many creators say they receive: an automated system has determined their content violates policies relating to nudity or sexual content.
It is becoming an increasingly familiar story for LGBTQ+ creators. Years of work disappear overnight. Entire YouTube channels vanish without warning. Videos are removed. Monetization is stripped away. Appeals are rejected or met with silence. Then comes the same explanation many creators say they receive: an automated system has determined their content violates policies relating to nudity or sexual content.
For
those whose livelihoods depend on digital platforms, this is not merely
an inconvenience. It can mean the loss of years of creative work,
thousands of subscribers, advertising revenue, sponsorship opportunities
and, in some cases, a significant portion of their income. A channel is
not simply a social media account. For many, it is their business.
To
be clear, there is no credible evidence yet that YouTube is
deliberately conducting a coordinated campaign to remove LGBTQ+ creators
simply because they are LGBTQ+.
However, that does not diminish the
seriousness of what many creators say they are recently experiencing.
Reports of wrongful strikes, demonetization and channel removals have
become increasingly common, and many affected creators believe automated
moderation systems are incorrectly flagging perfectly legitimate LGBTQ+
content as sexual or otherwise inappropriate.
That
distinction matters. Intent may not be discriminatory, but outcomes can
still disproportionately harm a particular community.
The
timing is difficult to ignore. YouTube has significantly expanded
automated enforcement across the platform, relying more heavily on
artificial intelligence to detect potential violations before human
review. While creators across many genres have reported erroneous
strikes, LGBTQ+ creators have become particularly vocal about what they
see as algorithms that fail to distinguish between sexual content and
LGBTQ+ identity or expression.
Travel creators
documenting Pride festivals. Couples sharing their daily lives. Gay men
enjoying a beach holiday. Educational discussions about LGBTQ+ issues.
Lifestyle content. None of these are inherently sexual. Yet many
creators claim their content is repeatedly caught in moderation systems
that appear unable to make that distinction.
This is not an entirely new problem.
Back
in 2017, YouTube acknowledged that a technical issue with its
Restricted Mode incorrectly filtered hundreds of thousands of
LGBTQ+-related videos, effectively hiding them from many users. The
company eventually corrected the issue, but it demonstrated something
many LGBTQ+ creators had already suspected: algorithms can reinforce
bias even when discrimination is not intentional.
Nearly a decade later, many creators fear history is repeating itself.
The concerns extend beyond YouTube.
Across
Instagram, Facebook and other Meta-owned platforms, LGBTQ+ creators
have spent years battling opaque moderation systems that frequently
confuse identity with explicit content. Images of men at beaches,
shirtless photographs taken in perfectly ordinary settings, or
promotional material for LGBTQ+ venues can trigger removals or account
suspensions despite containing nothing that would reasonably be
considered pornographic.
My own experience reflects that frustration.
Over
more than a decade managing LGBTQ+ social media accounts, including Gay
French Riviera, I have seen my accounts suspended multiple times and
anywhere from 3 to 30 days with appeals going nowhere. But at least Meta
generally provides some explanation, even if that explanation often
appears disconnected from reality.
As an example,
a photograph of a man standing on a beach on the French Riviera somehow
becomes problematic despite beaches being full of people wearing
swimsuits every single day.
AI-generated promotional images featuring shirtless men—even when they are not touching or posed suggestively—have also triggered restrictions. Some LGBTQ+ venues have become almost impossible to promote because automated systems appear to treat their imagery differently than comparable mainstream tourism content.
AI-generated promotional images featuring shirtless men—even when they are not touching or posed suggestively—have also triggered restrictions. Some LGBTQ+ venues have become almost impossible to promote because automated systems appear to treat their imagery differently than comparable mainstream tourism content.
Even language
itself seems to create problems.
Many creators now warn each other
against combining perfectly ordinary words such as “travel,” “beach,”
“gay” and “men” in titles, descriptions or hashtags out of fear that
algorithms may incorrectly classify their content.
Whether
every one of these claims can be independently verified is beside the
point. The fact that creators are modifying perfectly legitimate speech
to avoid invisible moderation systems demonstrates how deeply trust has
eroded.
Adding insult to injury is the lack of transparency.
When
a creator spends years building an audience, carefully following
platform guidelines, investing countless hours filming, editing and
uploading videos, they deserve more than an automated email generated by
artificial intelligence. They deserve to know precisely what policy was
violated, exactly which content triggered enforcement and the
opportunity for meaningful human review.
Instead,
many describe appeal systems that feel equally automated, leaving
creators trapped in a bureaucratic loop where no one appears willing—or
able—to explain what actually happened.
Organizations
such as GLAAD have also criticized YouTube’s moderation practices,
arguing that recent policy changes have contributed to a less safe
online environment for LGBTQ+ users and creators. While GLAAD has not
alleged a deliberate purge of LGBTQ+ accounts, its concerns echo what
many creators have been saying for months: automated moderation systems
are producing outcomes that disproportionately affect LGBTQ+
communities.
Artificial intelligence undoubtedly
plays an important role in moderating billions of pieces of content
uploaded every day. No reasonable person expects every video to receive
immediate human review. But when AI systems repeatedly struggle to
distinguish between LGBTQ+ identity and sexual content, the technology
is no longer simply imperfect—it begins producing systemic consequences.
These are not isolated technical glitches. They become barriers to participation, visibility and economic opportunity.
For
LGBTQ+ creators, visibility has never been guaranteed. It has been
fought for over decades. Social media platforms often present themselves
as champions of diversity during Pride Month, wrapping logos in rainbow
colours and celebrating inclusion through carefully crafted marketing
campaigns.
Those commitments ring hollow if the very
systems powering these platforms continue to suppress, demonetize or
remove legitimate LGBTQ+ voices through flawed automated enforcement.
If
artificial intelligence is making these mistakes, then the companies
deploying that technology have a responsibility to fix them.
That
responsibility includes meaningful transparency, faster access to human
review, detailed explanations when enforcement occurs, independent
auditing of AI moderation systems for potential bias, and public
reporting on whether certain communities are disproportionately affected
by automated decisions.
Until that happens, every
unexplained suspension sends the same troubling message to LGBTQ+
creators: your years of work, your audience and your livelihood can
disappear at the click of an algorithm—with little explanation and even
less accountability.
Some affected creators have suggested filing complaints through the Digital Services Act framework in the European Union, including with their national Digital Services Coordinator, as one avenue for challenging moderation decisions where applicable. Whether that process proves effective remains to be seen.
What
is already clear is this: if Big Tech wants creators to trust
artificial intelligence with decisions that affect careers and
communities, then AI cannot become an unaccountable black box.
Technology should protect free expression, not repeatedly place
marginalized communities in the position of having to prove that simply
existing, travelling, speaking or sharing their lives is not a policy
violation.
This seems to be just one more battle for us to fight. But fortunately for us, this ain’t our first time at the rodeo.
