There is something wrong with the way we build large, consumer-focused technology companies.
Uber lost $21B in value because it lacked ethics and created a business by systematically breaking local law in its largest markets. Twitter helped elect a despot because they cared more about engagement numbers than building a safe product. Facebook shipped advertising software that was designed to be used as a tool for espionage against democracy and discrimination against people.
These are not aberrations. These are fundamental features of the companies that allowed their owners to build their wealth to our detriment. Our current system of making companies has lead to the creation of software that hurts us.
In order to fix that, we need to make fundamental changes in the way corporations work. We need a new model for corporate governance, one based on daily checks & balances rather than board control. A model based on stitching together a number of currently weak or non-existent corporate functions under the banner of what we’ve traditionally called Corporate Social Responsibility (CSR). All of this must be designed to ensure corporate profits in the context of stewardship & care for the people of Earth. There would be five functions under a renewed and strengthened CSR division:
- Customer service
- Quality assurance
- Public editor
- Inspector General
Here’s a schematic:
All of these roles exist, and many are already deftly implemented in tech companies. Nothing I am writing here is ground-breaking or new— I just wanted to pull it all together into one view and use it as a way to organize and document how companies adopt these practices.
In other words, the future of ethical corporate governance is clearly already here, it’s just unevenly distributed. Here’s a look at each component:
The customer service function in major tech companies has largely developed around issue tracking. The idea is that when a user seeks to communicate with you outside of the bounds of your interface, you have to correct that behavior by getting them back into the interface. This is the opposite of active listening. The typical issues are inability to log in, or they’ve discovered a software bug, or they simply haven’t read the manual.
Smart tech companies are good at using data generated by issue tracking systems to make their software better. But using big data to make incremental improvements to existing code can obscure the need to reconsider how what you’ve built actually serves humanity.
This quantitative data has to be considered along with qualitative information about how people feel when they approach you. We need a powerful customer service function to constantly be the voice of and advocate for people. This happened, informally, at Facebook after the 2016 election, when engineers staged a mini-revolt to force the company to consider how it was used to advance treachery. We need make that normal and expected.
QA is another long-standing function in tech companies that should be lifted up and revamped.
In general, the Quality Assurance function is not held in high regard in a technology company. The people who test software before it is launched are traditionally paid lower and have less status than the people who make software. That disparity is a central feature of techbro arrogance.
The scope of testing in QA is far too narrow to meet the needs of society. “Test cases”— the scripts used to test software— are typically derived solely from the software specification— the document that supposedly guides development. Neither internal conditions— “is the developer of this software biased against certain users?” nor external conditions— “does this software break laws in places where it is used?” are considered.
We should set QA free. Instead of “can a user in Pittsburgh indicate their location and secure a ride”, it should be, “can a user in Pittsburgh indicate their location and secure a ride without breaking the laws of Pittsburgh”. In order to do this, the QA people have to know the law, and write cases that test that. Larry Lessig clearly laid this out almost 20 years ago in Code and Other Laws of Cyberspace. The work of Carl Malamud is instructive as well— indexing and making our laws widely available is a precursor to embedding law into our technology systems.
When ProPublica ran a series of articles on how the Facebook advertising platform allows people to systematically break federal discrimination law, all they were doing was QA. This function was escaped from the corporation; it should be inside it.
The public editor is a role from the journalism industry that is sorely needed at any company that maintains a platform for content. The job, in news, revolves around the implementation of proper journalism ethics, identifying and examining critical errors or omissions, and acting as a liaison to the public.
In tech companies, this role is again broken into separate pieces, but mainly delegated to automated customer service systems for flagging posts. Systems like this proudly prioritize efficiency over empathy. We need a human with the power and the charge to make sure the content on the platform helps humanity.
The lobbyist function has always been the most aggressive of this bunch when it comes to protecting corporate rights, and I think that should remain the case. There’s nothing wrong with a zealous corporate actor that puts the health of profits before other interests. What’s often missing is the counterpoints to that zealousness that places corporate objectives in the proper context of sound public policy, informed by the other work of other offices that help in accountability.
I remember after the 2012 Obama campaign and people went on to serve as lobbyists at Uber and other companies. As far as all can tell, all they did was continue the same patterns of protection and insider mojo. I find it distasteful.
The foundation of accountability in a corporation lies with the board of directors. Having an office of inspector general, with independent capacity to investigate corporate wrongdoing and report it to the board, would go a long way toward keeping corporate actors honest.
The history of offices of inspector general closely patterns after major scandals that reduce trust in institutions. The New York City Department of Investigation— widely considered the cream of the crop— was founded in 1873 after the Boss Tweed and Tammany Hall scandals. The first federal non-military Office of Inspector General was established in 1976 under the Department of Health and Human Services to fight waste, fraud and abuse in Medicare, Medicaid, and more than 100 other HHS programs. In 2005 I helped launch a website for the new Chicago Inspector General, which was strengthened after the Hired Truck Scandal.
Let’s do this
Systemic trauma like what we’ve experienced in our online platforms requires systemic remedies. And when corporate actors give us no reason to trust them, we need to add new methods of accountability to restore trust. The more that private corporations become instruments of our collective civic voice, the more it is necessary that we inject government systems of accountability into how they marshal our data.
If this isn’t Corporate Social Responsibility, I don’t know what is. Far more than a Saturday morning painting fences at the local park.