Understanding harassment on metaverse platforms, with suggested actions

Harassment in the metaverse has been an ebbing topic for some time, sparked by mainstream media outlets covering cases in early 2022. I've been following this for some time, from research papers to reading about Nina Jane Patel's experiences. But the final spark that brought pen to paper was, rather typically, an abrasive LinkedIn conversation. The person did not think harassment in the metaverse was 'real,' because it was all virtual. He argued that touch and proximity meant nothing in a virtual space, with flippant dismissal.

This rubbed me the wrong way; harassment takes many forms, and physicality is not a key differentiator. It was like speaking to someone from the 1990s who did not believe internet bullying can exist. A study by Timothy Taylor articulated it well:

It's a key question too, when harassment is so common worldwide. A European Union agency found that one in four Europeans were victims of harassment (in some form) within one year. Similar findings were found in the US, where women were harassed persistently in online spaces. (While harassment occurs regardless of gender, the vast majority of victims are women. Plus, McKinsey found that more women engage on metaverse platforms than men, which may make the problem more acute in virtual spaces).

We know that it's an understudied form of harassment that is emerging today, which has been underpinned by numerous studies. What's less clear is the extent to which harassment is common in the metaverse, and how its unique nature requires a different approach to combatting the area. The article pieces together what we know, what can be done, and where we need further work.

Some dismiss harassment in the metaverse as niche and superfluous, shrugging off its impact. I disagree. We should lay out a framework and approach to minimise harassment, so that we can reduce widespread harm as the metaverse grows.

Executive summary

• Lack of research: We have few good studies on harassment in the metaverse, where it is ripe for an expansive survey among users across multiple (or individual) metaverse platforms. Most research is qualitative, and not enough is quantitative.

• Good progress to empower users. Metaverse platforms are implementing updates so that users can protect themselves. Good progress has been made in this area over the last year.

• Education as an area of improvement. A key issue is that many harassers are either oblivious to their harm, or are unaware of its impact. Metaverse platforms could take a stronger role in educating users.

How much harassment is there in the metaverse?

It is hard to say. We have scant research on the subject, and the few reports we have are riddled with caveats that complicate the picture further. The most robust study comes from a Swiss anthropologist and Japanese VTuber, who received answers from several social VR users and analysed the data.

First, some caveats. The survey gathered responses from ~900 people who use social VR applications regularly - not that many when it comes to a widespread study. The vast majority of the responses (743) came from Japan, so I would treat this survey as a Japan-focused study rather than a global one. Also, the majority uses VRChat the most (648 of responses), which narrows it further to the experiences of VRChat users.

With that said, it does show some disappointing trends. Half of the users have experienced harassment in some form or another, with sexual language be using used the most. Unwanted touching and entering physical spaces were the top forms of harassment as well.

The insights are not surprising, but may not be representative of the wider field. The difficulty is data. Finding multiple social VR users for a survey is difficult, and studies like these are important - even though it's the biggest one with social VR users I have seen. I would like to see a more representative sample someday, perhaps focusing on countries or other platforms. From there, it will be far easier to see what needs to be built or addressed to protect people.

How to protect people in the metaverse

How can users protect themselves? I see it as three approaches:

• Empowerment via tools: Equip users with the tools necessary to ensure that bad actors cannot step into their private spaces. This can take the form of blocking people easily and seamlessly.

• Prevention via education: People who use metaverse platforms should understand the new ways in which they could harm others. This can take the form of articles, nudges online, or consistent communication.

• Elimination via administration: Removing bad actors who break the policies of metaverse platforms by harming others.

I will not touch on the third point as much, as most metaverse platforms already have policies in place to remove users. The same goes for companies that manage a corporate metaverse; harassment principles in the workplace apply in online spaces, too. The effectiveness of removing users might be interesting, but it's hard to say without knowing how much harassment there is across multiple platforms. More research is needed to understand the effectiveness of elimination.

VR in particular presents difficulties in monitoring activities on harassment. One paper suggested a process of 'responsive regulation' that offers penalties that are proportionate to the wider context. The process would 'empower communities to take swift action against violators, without alienating users for incorrect enforcement decisions.' Think of it as flexible administration from the bottom-up, which web3 enthusiasts may like.

Beyond frameworks, it's also important to conduct the risk assessment as well as possible. 'It is unrealistic to release the responsibility of preventing harassment to the potential victims, as that is overlooking the natural human stress response which often prevents the individual being attacked from taking action,' said Marion Ratié, Senior Digital Producer at Landor & Fitch. 'As metaverse builders are creating new spaces of interpersonal interactions, the basic mechanics of those must be researched with as much effort as the technology that’s being developed. The associated subject matter experts should be represented as much as the strategists, creatives, technologists and developers who make metaverse project teams.'

With that, let's explore the other two approaches.

Empowerment via tools

Different metaverse platforms have integrated differing layers in which people can protect themselves. The most publicised came from Meta, which implemented a Personal Boundary to Horizon Worlds. Users can stop who can access their personal space, based on their own comfort. 'We believe Personal Boundary is a powerful example of how VR has the potential to help people interact comfortably. It’s an important step, and there’s still much more work to be done,' the company said when they announced the new feature.

Harassment also occurs in platforms like Decentraland, The Sandbox and Roblox, and all have ways to report users who do harm. Blocking is the most common, and is widely used among players. Companies can also step in if there are reports - though the same tools can also be used to flag false reports as well. One Roblox player found a group who 'spam reported' them consistently. But overall, blocking tends to work well for most.

What other tools might work in the metaverse? For social VR platforms, a proximity tool works. Perhaps another is to ensure that users cannot enter virtual worlds, so that private spaces can be protected. Better administration tools for players would help to empower their own protection, both around their bodies or in virtual locations. One victim I spoke to also recommended a 'fast exit' button, so that they can leave an uncomfortable situation swiftly. The better the tools, the more users can take hold of their own fate.

One counter-argument is that it invokes the idea that people are 'property' that portrays bodies as trespassable. 'This is similar to discourses that hold women responsible for communicating (non)consent, position women as the gatekeepers of sex, which ultimately work to sustain the rape culture,' argued Trang Le from Monash University. She went on to argue that we need to address 'structural and systemic conditions' alongside technological changes - which I agree with wholeheartedly.

Prevention via education

Education is a key one for me. My interaction on LinkedIn showed me that many people do not believe harassment in the metaverse is 'real,' which can be a dangerous path to go down. Plus, not all harassment is malicious at heart. Much of it can come from oblivious people who harm others accidentally, due to a lack of understanding or cultural differences. Unintentional harassment is already codified in institutions like the UK Parliament.

That is where education comes in. I believe that it is important for people to fully understand what constitutes harassment in the metaverse, cementing how realism is not a prerequisite for harm. The onus falls on metaverse platforms to an extent, perhaps with tooltips on why tools can be used or via updates across communication channels. The same responsibility falls on companies as well, to administer HR policies with the same rigour as offline spaces.

One difficulty is that harassment education may not work as well as other methods. A study that focused on higher education found that even in-depth training lacks genuine impact. This needs a pinch of salt, as it's difficult to equate findings from one area to metaverse platforms. Yet even so, some level of information given to users would not go amiss. Better some than none.

Conclusion: How to address harassment in the metaverse

Based on speaking to victims, it is clear that a dismissive attitude permeates across metaverse platforms. It has an impact on virtual spaces and beyond, affecting how people converse and attend events. One study found that attendees to VR concerts fear harassment when they attend - a layover from both real-life and virtual experiences. A safe space is for the benefit of everyone.

The lack of understanding is natural while the metaverse is at such an early stage, and platforms are already taking steps to empower users. But I believe that some level of education by platforms and businesses may help to cut attitudes before they embed into the new digital future, well before the weeds take root.

I also believe that we need more research as well, beyond surveys on attitudes and smaller ones colloquially done by enthusiasts. Robust work needs to be done to understand the area, and then use the information not design ways to better protect users. It is unclear how expansive harassment is; but with troubling insights like the below, it leaves a real impact:

It's a hard task. Harassment goes beyond online spaces and permeates around the world, impinging people's personal spaces and harming many people. The problem steps beyond simple technological fixes or blogs. It's a structural problem, and bold steps in the real world may seep into virtual ones as well.

One step in the right direction is recruitment. 'I believe a major reason for that is the lack of representation of the population at risk themselves; women are currently underrepresented in the growing metaverse industry,' said Ratié. 'It’s an issue that could be tackled within the company’s recruitment policy itself. If that population was represented in the making of those projects and products, the chance of tackling the issue at the source would be increased. Representation is key to an inclusive product.'

Tacking harassment in a comparatively small area may lack a wider impact. But a real one is happening, and severing the issue soon via structural reforms may succeed. But the first step comes from a solid understanding via good research. While we have a lot of qualitative studies, we need more data to take affirmative action.