Thread by @The_Crox

Under the skin of #OnlyFans: From receiving terrorism videos to hosting revenge porn, my latest investigation based on the experiences of dozens of women reveals concerns about how the British-run site is structured, managed and moderated. I found...
bbc.co.uk/news/uk-572699…

IS terror videos, rape threats and racism has been posted on the site in the past year in violation of its terms of service, as well as explicit photos of a student uploaded and sold without her consent for nearly four months. OnlyFans bans revenge porn and requires any users...

making content featuring other people to have written proof of consent, or says they must also be registered creators. But I understand proof of consent is only requested if an account is flagged or reported, which relies on people being aware their images are behind the paywall.

One creator, Victoria-May, was told "rude messages" did not violate OnlyFans' guidelines after she reported rape and death threats, and the subscribers' account would not be removed. The site then took six months to respond to more than 20 successive complaints...

But as the company takes 20% of all creator earnings, she feels it has a duty to take action. OnlyFans says it handled her complaints "in a timely manner" and that any fan reported and found to have been abusive is immediately removed from the platform.

Data protection experts have raised concerns about the security of OnlyFans' server after creators were hacked or affected by glitches resulting in pics of others appearing on their accounts - incl one who had an image of a man holding his penis appear instead - which OF refutes.

Tina Bean was living in her car before she started making $2,000 a month on the site. But this changed overnight in January after a hacker seized control of her account, blackmailed her for $150, uploaded streams of IS terror videos, stole all her photos, and posted racial abuse.

OnlyFans said her account didn't have two-factor authentication, which made it vulnerable. It says it's moderation system has now been upgraded and, "in light of this particular incident", protocols updated to ensure any hacked account is fully reviewed before access is restored.

But another creator, Katja Miyatovich, was hacked in May despite having two-factor authentication security enabled to prevent this. Data protection experts say the glitches and hackings should've been reported to the Information Commissioner's Office and 2fa should be mandatory.

Creators also said the policy of granting automatic refunds to blocked fans can perpetuate and incentivise abuse. OnlyFans has now adjusted its processes around closing creators' accounts after I raised why one had her page shut without explanation, leading to it being restored.

"We're the ones putting the food on their tables. So, when I'm being harassed and insulted, I'd expect [OF] to allow me to block someone without them getting a refund," says SkylaRayne. But creators say blocked fans can still create new accounts and sign-up to their pages again.

OnlyFans explicitly states that it's "not responsible" for moderating content, or detecting breaches of the law or its policies. But it told us that, while it was not legally obliged to proactively do so, it goes "above and beyond" what current regulations require....

by using community reporting and artificial intelligence software to help keep users safe. The site says it evaluates over 300,000 media files a day, and has more than 500 agents involved in compliance and moderation, with all flagged content reviewed on a case-by-case basis...

But this could amount to at least 600 images per person with some videos alone lasting a 1hr+. Digital consultants say artificial intelligence can be unreliable for porn, and OnlyFans needs more human moderators as explicit content is more susceptible to abuse or copyright...

OnlyFans said, as a subscription-based platform where fans pay for access and content, it was "uniquely geared to dissuade and deter abusive behaviour from users" because, unlike anonymous social media, it could swiftly identify and remove them.