Whether it is fake reviews that mislead consumers and harm businesses, fake user activity that defrauds advertisers, or fake news that is used to manipulate elections and markets, fakery is pervasive on the internet. The scale is massive and the economic and social impacts are profound.
Fakery and Fraud on a Massive Scale
Jupiter Research estimates that marketing fraud will cost advertisers $19 Billion Dollars in 2018. (https://www.raconteur.net/risk-management/tackle-marketing-fraud)
It is estimated that over half of all web traffic is from bots. (https://www.theatlantic.com/technology/archive/2017/01/bots-bots-bots/515043/)
Facebook detected and deleted 583 Million fake accounts in just the first quarter of 2018, but they estimate that Tens of Millions remain. (https://www.cnet.com/news/facebook-deleted-583-million-fake-accounts-in-the-first-three-months-of-2018/)
Twitter recently deleted Tens of Millions of fake accounts, roughly 6% of their user base, in what is only their most recent clean up of fraudulent accounts. (https://www.nytimes.com/2018/07/11/technology/twitter-fake-followers.html)
The Washington Post found that for some popular categories of products on Amazon the vast majority of reviews are fraudulent.(https://www.washingtonpost.com/business/economy/how-merchants-secretly-use-facebook-to-flood-amazon-with-fake-reviews/2018/04/23/5dad1e30-4392-11e8-8569-26fda6b404c7_story.html)
And of course there are the infamous Russian Bots. (https://www.salon.com/2018/07/09/russian-bots-are-back-walkaway-attack-on-democrats-is-a-likely-kremlin-operation/) (https://www.nbcnews.com/tech/social-media/russian-trolls-pushed-graphic-racist-tweets-american-voters-n823001)
Fake and fraudulent activity on the internet is resulting in direct losses of Tens of Billions of Dollars per year. There are Hundreds of Billions of Dollars of purchasing decisions per year being manipulated by fake reviews. Both businesses and consumers pay the price for all this fraud and the magnitude is so great that ultimately the entire economy is impacted.
The social costs of illegal political propaganda, or what might be called information warfare, are impossible to quantify but the threat that this activity poses is even greater than that of simple profit motivated fraud.
Digital Real IDs
The concept of a Digital Real ID is simple: in order to make a post on Facebook, write a review on Amazon, log in to your bank account, or send an email you have to authenticate yourself with an ID that is registered with your real name and issued through a secure process equivalent in nature to the issue of a Driver’s License or State Identification Card.
In practice there are two deal-breaking objections to Digital Real IDs:
- Such a system is technically impossible to implement
- Anonymity is a basic right that must not be infringed
If it is assumed that anonymity or pseudonymity must be part of the system, to the extent that these are protected legal rights, the only real challenge is to design a Digital Real ID system that actually works, while still allowing for anonymity in the contexts where it is a protected right.
Many people say that such a system is impossible, but that answer is simply not acceptable in the face of the enormous economic costs and threat to civil society posed by digital fakery and fraud.
A Digital Real ID system that protects anonymity is not only possible but necessary.
Building the Foundation
The foundation of a Digital Real ID system must be built into computers, mobile devices, and network infrastructure at the hardware level.
Each Digital ID must be implemented as a physical device, similar to a SIM card, or USB dongle, that can be easily swapped between different devices.
Each Digital ID device incorporates a simple computer processor that is limited to receiving a message from the host device, signing that message with a digital signature, and then sending the message back to the host device.
Each Digital ID device has a unique digital signature that is implemented in the device hardware and is impossible to change.
Every device that connects to the internet must have a network chip that requires a Digital ID device to be present in order to operate and every single packet of data sent by the network chip must be digitally signed by the Digital ID device.
Networking infrastructure hardware must be designed to read and verify the digital signatures on each data packet and must prevent any data packet without a valid digital signature from being transmitted.
With this foundation in place every single packet of data that transits the internet can be positively and indisputably associated with a particular Digital ID.
A Digital ID by itself is still anonymous or pseudonymous. In order to turn an anonymous Digital ID into a Digital Real ID rules, processes, and systems to associated Digital IDs with Real IDs must be established.
Managing Digital Real IDs
The hardware infrastructure that associates data sent over the internet to individual Digital IDs is independent from the system that associates those Digital IDs with Real IDs. This allows the same hardware to be used globally while allowing local jurisdictions to establish their own Real ID registries and set their own rules for what types of online activity require Real IDs.
While the details can vary between jurisdictions each jurisdiction maintains its own registry associating Digital IDs with Real IDs and/or maintains a list of registries that are trusted and used.
A Real ID registry would allow individuals to establish a Real ID by providing an existing government issued ID in person and associating that Real ID with one or more Digital IDs.
As long as an individual is using an authorized Digital ID device they can add and remove Digital IDs from their Real ID profile online.
If a Real ID profile is compromised or an individual loses all of their Digital ID devices then they would need to go to a registry in person to re-establish access to their Real ID profile and register a new Digital ID.
Requiring Digital Real IDs
Digital Real IDs could be required at the network level in certain cases to completely prevent unverified data from transiting secure networks. In other cases Digital Real IDs could be required at the application level.
At the application level a Digital Real ID would work similar to existing third party authentication systems such as the buttons that allow you to sign in to an account with Google or Facebook.
At the application level a website would send a token to the Digital ID device and the Digital ID device would return this token with a digital signature. The application would then contact the Real ID registry and present the signed token in order to get the user’s Real ID.
The use of Digital IDs and Digital Real IDs would provide a single universal login system that would eliminate the need for users to manage usernames, passwords, and personal profiles for individual web sites and applications.
What Problems do Digital Real IDs Solve?
Requiring Digital Real IDs for social media posts and product reviews would eliminate the vast majority of fakery and fraud currently taking place online whether that be for financial gain or political purposes.
Even if a service does not require Real IDs, Digital IDs could still be leveraged to reduce fraud by introducing significantly higher costs for creating unique online accounts.
Marketing fraud relies on fake user activity, which must be faked very cheaply, so introducing Digital IDs that cost even a few dollars each would make most fraudulent schemes unprofitable.
Economically motivated fraud can always be eliminated by making it unprofitable and so the cost of Digital IDs can be adjusted through taxation to whatever level is needed to make the majority of fraud unprofitable.
Digital Real IDs will dramatically improve security for users by locking down their online accounts. Unlike existing two-factor authentication systems, Digital Real IDs provide a secure mechanism to recover access to accounts when Digital ID devices are lost or stolen.
The use of Digital IDs at the network level gives users the ability to manage their network traffic in a new way by setting rules for what domains or IP addresses their Digital ID device will issue signatures for. Through this mechanism Digital ID devices can act as an out-of-band firewall that cannot be hacked or bypassed even when used with compromised host hardware.
A Digital ID device is very simple and thus much easier to audit and secure than a modern computer or mobile device, which is essentially impossible to audit or secure. Requiring all internet traffic to be signed by a Digital ID device would put users back in control of their own computers and mobile devices by giving them the ability to monitor and optionally block every packet of data sent by their device over the network.
In the event that a Digital ID device was bypassed, such as by a compromised network chip, a user operating their own network could trivially detect and block such compromises by using networking hardware that only allowed traffic signed by an allowed set of Digital IDs.
Roadmap for Digital Real IDs
Building the infrastructure foundation for Digital IDs could be done by the private sector on their own initiative but it would progress much faster if a major government mandated it.
Realistically, either the U.S., E.U. or China would need to develop the hardware specification and mandate it for hardware sold in their market in order for implementation to be insured.
Once one major market makes Digital IDs a requirement it would probably be implemented in all new hardware and the private sector in other markets would include it voluntarily.
Requiring Digital IDs at the internet backbone level, setting up Digital Real ID registries, and requiring certain online services to use Digital Real IDs can all be done incrementally on a jurisdiction-by-jurisdiction basis once Digital ID support is baked into new hardware.
Everyone stands to benefit from the use of Digital IDs and Digital Real IDs including individuals, businesses and governments. Digital IDs can be leveraged to eliminate Tens of Billions of Dollars in yearly fraud. Digital Real IDs can shift social media from being an existential threat to civil society to being the powerful enabler of civil society that technologists had hoped and dreamed it would be.
Anonymity or pseudonymity can be maintained as rights with the long established limitation that those rights do not extend to committing crimes or harming others. These rights will be subject to a small tax, which prevents them from being abused for petty fraud, but that is an entirely reasonable price to pay.
After many years of failing to act it is time for tech companies and leaders to stop claiming that Digital Real IDs are impossible or infringe on individual rights and start getting to work on implementing them. Ultimately, changes in law are needed, but the private sector will benefit greatly from being the leader rather than the follower in this process.
Stay tuned for part two: Strong End-to-End Encryption with Authorized Access for Law Enforcement is Possible and Necessary