Facebook's decision to remove videos showing people being decapitated leaves the firm in a quandary: should or shouldn't it impose a wider censorship policy?
Originally the social network rejected calls from users to delete the clips saying that it wanted to "preserve people's rights to describe, depict and comment on the world".
But after the BBC revealed that one of its own safety advisers - the head of the Family Online Safety Institute - had criticised its decision, the firm announced a U-turn, saying it would remove clips showing beheadings while it re-evaluated its rules.
That potentially opens a can of worms.
Since publishing the article, readers have contacted the BBC to complain about other videos, including:
- one that shows killings which do not involve beheadings
- clips involving cruelty to dogs and other animals
- a smartphone recording of a schoolgirl being punched to the ground by another pupil
In all cases they said the network had refused their requests to remove the material. A spokeswoman for Facebook confirmed its policy had only been amended in regard to decapitations.
But imposing stricter controls would open the firm up to other criticism.
Before his death, internet freedom campaigner Aaron Swartz warned of the dangers of privately owned parts of the net limiting what was posted onto their sites. He called this "corporate tyranny" and named Facebook as a specific concern.
The social network could not provide a date for when its review would be complete. The following range of opinions suggest it will struggle to please everyone.
Richard Allan, Facebook
More than a billion people express themselves and comment on the world in which we live through Facebook and most of the time this is entirely without problem.
On occasions, there are concerns about some of the content that is being shared and we have put in place a reporting system so that people can tell us about this.
The reported content is evaluated against our community standards and appropriate action is taken where our rules have been breached.
When drawing up and enforcing our approach to acceptable behaviour and content on Facebook, we aim to strike the right balance between enabling people to share information, news and content - and protecting the community as a whole.
This is a complex challenge as Facebook is a large, diverse community and we are continually presented with novel situations.
While we freely admit that we do not always get it right, the trouble-free daily experience of the vast majority of Facebook users demonstrates that our systems are working well in all but the most exceptional cases and that they are improving over time.
As we said last week, we are reviewing our rules related to content showing graphic violence.
In doing so we are clear that there are situations where it is important for people to be able to share content through Facebook even if this can at times be quite shocking.
For example, people caught up in violent incidents such as the recent Boston bombings or the ongoing conflict in Syria want to be able to report on their experiences and may use quite graphic content to do this.
This illustrates the kind of challenge that our highly experienced team deals with on a daily basis as we strive to offer a space for sharing that is mindful of everyone's expectations.
Celia Mellow, petition organiser
As a person who holds a strong sense of justice, I had no hesitation in setting up a petition for the removal of the sickening decapitation video I was shocked to find on my Facebook news feed.
What shocked me even more was the fact that I had to actually make a petition in any hope for the video to be removed.
No matter how many times my friends and I reported it, we all received the same message, stating that "it doesn't violate Facebook's community standard on graphic violence, which includes depicting harm to someone or something".
How does a video of an innocent woman being brutally murdered not "violate" this? I can only hope that there is a criminal investigation that will bring her justice.
As a loyal Facebook fan, I understand that Facebook is only allowing people to have freedom of speech. However, I think it is about time they drew a line between what is and isn't appropriate for the public.
Facebook's audience starts from children aged 13 - what I feared the most was that my younger sister could easily have witnessed that disgusting video.
No-one should be exposed to such graphic horror. Sadly, that video isn't the only inappropriate content to have wandered onto Facebook recently. I have heard of others showing extreme violence and cruelty to both humans and animals.
It's time that new stricter regulations are made by Facebook in order to remove these vile videos for good so that it might return to being the safe social network it used to be.
Jeremie Zimmermann, La Quadrature du Net
Any intervention by Facebook to remove or block access to content beyond what a court might order - while respecting basic fundamental rights and the principle of proportionality - would in practice amount to privatised censorship, and nobody has an interest in going there.
A dominant, centralised actor such as Facebook would be incentivised to spend as little money as possible determining which content would be lawful or not, suitable or not, etc.
This would raise the question of what criteria would be used. Opening such a breach would ensure that any government could pressure Facebook to consider their own criteria, whether for political, religious or other reasons.
Under such conditions we can be sure that the fundamental right to freedom of speech or the right to a fair trial would not be respected.
As surely as we cannot trust giant centralised corporations to defend our fundamental freedoms, we cannot ask them to become the judges and enforcers of what information should be shared online.
Protecting children on the net is a responsibility of their parents in the first place. It cannot be outsourced to Facebook.
It is a matter of educating them about the difference between between privacy, publicity and a circle of trust.
Since Facebook collects and stores so much information it should be able to determine when one of its members is a minor and is about to be exposed to content that has been reported as unsuitable, and display a warning message.
Users would then be free to choose to take that advice, or make a conscious choice to access the content.
Stephen Balkam, Family Online Safety Institute (Fosi)
Facebook, and most other social media sites, have explicit terms of service about what is and what is not acceptable to be hosted on their websites.
Some go further and have created what are known as community standards.
These more clearly state the rules about what kinds of content will be removed.
Facebook, YouTube and Twitter have robust reporting mechanisms so that ordinary users can flag inappropriate or abusive content for review.
What is challenging for these companies is how and where to draw the line.
This will help them determine when to invoke the "public interest" principle in keeping material - such as images from the Boston Marathon bombing - up on their site, even though they depict graphic violence.
This is new territory for us all as we navigate the rules, ethics and standards of user-generated content sites.
Andrew McDiarmid, Center for Democracy & Technology
The controversy over Facebook's treatment of shocking videos of beheadings is the latest illustration of the enormous complexity at work when it comes to promoting the exercise of human rights online.
Billions of people rely on internet platforms to speak and access information in the networked public sphere, but the platforms are controlled by private companies, whose terms of service in large part determine the contours of free expression.
In one sense, platform operators are themselves speakers that have the right to determine their own policies. At the same time, these "digital sovereigns" - to borrow a phrase from Rebecca MacKinnon - effectively govern their users' exercise of free expression rights.
Platforms have a responsibility, particularly as they grow to Facebook-scale, to consider the human rights impact of their policies and to minimise restrictions on free expression.
This is especially true with respect to government restrictions. It would be troubling indeed if government pressure precipitated the video's removal in this case.
A key step in carrying out this responsibility is ensuring that content policies are clearly communicated and fairly applied.
The horrific beheading video and Facebook's reported reaction demonstrate the challenges that arise when trying to develop and apply clear, consistent standards in the complex and multi-faceted realm of online communication.
Context matters a great deal. Different companies might draw the line in different places, and just because something is offensive or disturbing does not mean it necessarily violates a particular term. And it certainly does not make it illegal.
Because of this complexity, systems for assessing content require constant refinement to ensure that free expression is protected.
Advocates, too, must remain vigilant that the private players that provide so much public value online are meeting their responsibilities to users.
Is it complicated and prone to mistakes and close calls? Yes, but the alternative - mandated content policies and individual governments vying for control over the global internet - is untenable and fraught with risk for free expression.
Dr Lynne Jordan, British Psychological Society
The main concern, as an experienced psychologist, in working with the effects of actual and vicarious violence is a lack of awareness of violation of choice.
Material is posted on news feeds and "liked" indiscriminately without thought as to the rights of under-aged youngsters and others who may view it.
People, whether young or old, can be negatively affected by witnessing violence either on screen or in reality.
Effects include trauma responses such as replaying the images, feeling scared and vulnerable, ashamed, invaded or violated and confused, as well as angry and helpless, which is reinforced via the news feed as these things pop up uninvited.
Ethical codes are there for safety and to preserve the right to choose what is viewed when users are considered of age or able to understand the implications. Social media sites are mostly not obliged to adhere to such codes which creates a problem, particularly if they issue their own vague inadequate guidelines.
Social networks' news feeds allow material to arrive on people's pages that might never be sought by choice.
Extensive "friendship lists" develop with people who may not be actual friends but through casual contact get "befriended", perhaps out of obligation or a need to fit in, be liked etc.
The material is often posted supposedly to prevent the spread of violent crime or other violations, but in fact it can inadvertently escalate it by sidestepping the consent of the people accessing the feeds.
This is reminiscent of the "ban smoking in public places" debate with the concern of whose rights we are protecting.
In that debate it was largely about public physical health. This debate concerns public mental health and wellbeing.