About FLT Verified Badges

The Verified Human Badge is a pilot project by Field Level Tech. It is intended to demonstrate the potential of user verification to reduce the misuse of anonymous and fake social media accounts. It is informed by the work of independent NGO Clean Up The Internet, which has been conducting research and campaigning work on this issue for a number of years.

Verified Human badges offer BlueSky users an option to demonstrate to other users that their account is operated by a real-world human being and linked to their real-world identity. To gain a Verified Human badge on their user profile, a user must go through a verification process which requires them to demonstrate a real-world identity.

Verified Human badges are intended to provide users with a new tool to reduce the risks associated with anonymous and fake accounts. Verified Human badges aim to address two key ways in which anonymous and pseudonymous accounts are linked to a higher risk of harmful behaviour:

Accountability

To gain the badge, a user must provide a link between their user profile and their real-world identity, which is retained by FLT for as long as the account keeps the badge and a further year should they choose to remove it. This means a Verified Human has voluntarily made their profile more traceable. This should reduce the risks of negative behaviour associated with the “online disinhibition effect”. It makes it harder to circumvent blocks by other users or platform bans. In the event of a criminal act, it makes legal repercussions much more likely. A Verified Human should be less likely to be abusive or threatening or otherwise criminal - but if they are, other users, platforms, and law enforcement will have more options for tackling them

Individuality

We aim to limit Verified Human badges to one per user. In this pilot this can be circumvented for example by a person with multiple UK bank accounts. But this is not a loophole which is easily exploited at scale by a fraud network or foreign influence operation. A Verified Human is less likely to be part of a “bot” network seeking to manipulate or scam other users.

Please note this is a pilot project which is live to enable testing and encourage discussion and feedback. It is not at present a fully-fledged safety feature and should not be treated as such.

This feature has been developed making use of BlueSky's open protocols, but so far without any direct engagement with the company. We hope to engage directly with them in the future, as there are additional functionalities we would love to explore with them.

Known Limitations

Names are not verified

Verified Human doesn't mean a user is using their real name This pilot only verifies the link between a user's account and a real world identity. It does not check their profile name, or any other claims on their profile, or any other content they post. A Verified Human could be using a pseudonym. There remains a risk that an account with a Verified Human badge could make false claims about their identity, or impersonate someone else. But the account should be a lot less likely to be part of a coordinated network of inauthentic accounts. And if a Verified Human is perpetrating harmful deception then at least other users, platforms, and law enforcement will have more options for tackling them.

Verified accounts could use bots to post

Verified Human doesn't mean an account isn't using automation or AI to post content. The badge confirms that a real-world human has demonstrated their presence behind the account, and provided a link to their real-world identity. It does not check how any content is generated or posted.

Verified accounts could be hacked

Verified Human doesn't mean an account hasn't been hacked FLT is confident in the security of our own products, and this feature has been designed in a way which minimises its vulnerabilities - for example by minimising the data we retain to an unique, but anonymous, identifier. However, there are definitely points in this process which could be vulnerable to hacks and lead to a verified human badge being obtained by a hacker for an account which isn't traceable. A hacked bank account could be used in the verification process. (although OneID have processes in place to flag accounts with known issues, so once this is detected and reported to us we will remove the badge). Perhaps the highest risk would be that an account which obtains a badge legitimately is subsequently hacked. We therefore encourage all users to set a strong password and active 2FA for their BlueSky account. And we offer a mechanism for users of our service to report any breaches and will suspend any badges where a breach is reported. In addition, we periodically require users to revalidate their verification in order to retain their badge.

Bad people could verify their accounts

Verified Human doesn't mean the human behind the account is a nice person A Verified Human badge helps address a couple of major risk factors associated with anonymous and fake social media accounts, by introducing greater traceability, and being inaccessible to bad actors seeking to create scalable networks of inauthentic profiles. However, it does not eliminate risk. Crucially, a significant amount of harmful behaviour online is done by real people, who aren't hiding behind anonymity. Verified Human badges won't prevent this behaviour - although traceability should act as a deterrent and should make it easier to take action against serious and serial offenders.

Functionality we would like to explore in the future

Verified profile name

As well as “Verified Human”, it should be possible using our approach and third party verifiers such as OneID to offer a “Verified name” label. The safety considerations around this are more complex. And the verification services charge fees for this level of validation.

Filters on non-verified accounts

A core part of Clean Up the Internet's proposal is that alongside users having options to verify their identity, they should also have options to manage their level of interaction from non-verified accounts. Once uptake of a badge like “Verified Human” has reached critical mass, users should be able to choose only to hear from verified humans. This does not sit easily on top of BlueSky's current approach to 3rd party moderation and would require changes to the bluesky apps user interface.

Integration with bans and suspensions to prevent “verified human” phoenix accounts

If a “verified human” account is subsequently banned by BlueSky, we would like to ensure that the unique ID associated with that account is not able to be used to gain a badge for a new account. However this feature would require FLT Verified to be run with a high certainty of reliable operation.

A choice of verification providers

At present Verified Human relies on the “ID Live” service provided by OneID, which relies on users holding a bank account and being able to authenticate with either their bank online service or mobile banking app. This will work for a lot of people in the UK, but not everyone. Accessibility would be improved by offering users more choice of providers and methods of verification. We would like to integrate with other providers accredited under the UK's Digital Identity & Attributes Trust Framework, and delivering verification to a similar Level of Assurance.