Your Digital Footprint Could Make You Unemployable
In China, the government will arrest and torture you for criticizing them:
[In May 2022], Xu Guang…was sentenced to four years in prison...after he demanded that the Chinese government acknowledge the Tiananmen Massacre and held a sign calling for redress at a local police station. Xu was reportedly tortured, shackled, and mistreated while in detention.
In the UK, they don’t have freedom of speech, and up to 30 people a day (meaning 12,000/year) are arrested for the crime of posting or sharing “grossly offensive” content online.
In the US, we have freedom of speech enshrined in law—you can’t be arrested for saying politically inconvenient or offensive speech—but that freedom doesn’t protect you from the decisions of private companies. So if you’re a dickhead online, you can still be prevented from being hired, or lose the job you currently have.
This got me thinking…so I asked Claude the following:
Please web scrape information about me, [first & last name]. Based on what you find, please summarize, from an employer’s perspective, any red flags that make me unemployable.
Claude tore me to shreds—no mercy. Now I’m going about wiping my entire digital footprint. Am I overreacting? Well…
There’s currently a lot of uncertainty surrounding the advances in AI:
We are starting to see pre-AGI systems shrink analyst classes, change personnel strategies, and trigger layoffs.
So in an increasingly competitive and globalized job market, recruiters may employ (if they’re not already) web scraping AI tools that build comprehensive profiles of job applicants in order to find reasons to eliminate them.
The reason this could be tricky for employers (but certainly not prohibited) is because of privacy laws. As of 2023, the California Consumer Privacy Act now requires employers to report which categories of personal information they’re collecting, and it grants job applicants the right to request what was found. But notice:
It doesn’t prevent companies from collecting and using personal data against job applicants. It simply requires the disclosure of collected information upon request.
So what are some legal ways companies could discriminate against applicants?
Let’s say the CEO of a company is a teetotaler. She could use AI tools to surveil applicants’ online presence (including social media) and eliminate them if they’ve ever posted any images of alcohol, stating: “data collection uncovered drug-use that’s incompatible with the company’s values.”
With increasing political polarization, it’s important to remember that party affiliation is not a protected characteristic. Employers can absolutely discriminate against you for having specific political beliefs, which many people openly share on their social media.
These two examples presuppose that humans will always be the ones initiating employment discrimination. But with AI increasingly taking an active role in hiring, algorithms could make decisions on their own without humans ever directly being involved.
Continuing the alcohol example, an AI algorithm could examine the damaging effects of alcohol (on both an individual and societal level), and decide that alcohol probably makes people less productive employees. It could then automatically devalue job applicants that have ever posted images displaying alcohol (even if their photo was just from a common social function, like a wedding).
We’re moving towards a world where every publication of personal information can become permanent career collateral.
Maybe once I become financially independent (in a few decades), then I can begin posting under my own name. But for now, the risk of having a public online presence seems to be asymmetrically giving power to corporations at the expense of private individuals. Authenticity and personal accountability are becoming just another privilege of the rich.
Privacy law could catch up eventually, but Pandora’s Box has been opened: it’s now easier than ever to surveil and discriminate against job applicants. There’s now a strong incentive to anonymize your online presence in order to protect your economic future in a job market that has just obtained unbelievably advanced surveillance tools.
Sure, but she would probably go out of business unless she was operating in Saudi Arabia or Utah, compared to an equivalent company which hires everyone according to skill. This kind of arbitrary discrimination is so counter-productive that it’s actually immensely costly in secondary ways. In general, we should expect free markets to get better over time at optimizing hiring for job performance. If you’re a low-value employee (at or close to minimum wage) or if you live in a country where organizations are selected for non-market reasons (government cronyism, or something similar) then you’re not actually in a very free market so these things can still happen. Same for other cases of non-free markets.
Companies that discriminate make less money, but that doesn’t mean that the company goes out of business tomorrow. It just means that on the average, the company may not last quite as long as other companies. You still get a steady state where companies keep entering and leaving the market, but each company that enters with bad business practices lasts somewhat less time than the other companies. That will never result in such companies not existing in the market at all. (And woe be you if a company doesn’t discriminate, gets huge, and then starts discriminating. Its advantages from being big and entrenched are so big that discrimination won’t reduce its survivability by a noticeable amount.)
That also ignores the possibility of multipolar traps where most companies discriminate, and failing to do so gets a company in trouble from all the ones that do. They would be better off if nobody discriminated, but a single company who fails to discriminate would lose.
By your reasoning, Disney and Google would have gone out of business already.
(And I can name a lot of money-losing practices where companies with those practices are still around.)
I agree with you in principle, but I think we’re (in the U.S.A., at least) farther from a free market than most people believe. In particular, disparate impact can be claimed as a result of any policy change—whether to hire on a criteria or not to do so—and success or failure of these cases, accordingly, is determined primarily by precedent and the preferences of the ruling judge. Business owners I’ve spoken to have cited this as a reason every company in the country seems to have a very similar position on which non-protected characteristics are grounds for hiring/firing and which ones aren’t.
As an example, suppose I determine that workplace cohesion increases enough from a 100% teetotaler employee base that the smaller hiring pool is worth it. If someone I’ve ruled out takes issue, they can print out a set of statistics saying that protected demographic A is less represented among teetotalers than in the general population and that my alcohol-free workplace is “functionally discriminatory”, and they have a decent chance at winning a lawsuit on that basis.
You make a good point: advanced surveillance tools are more likely to be a problem for people living under autocracies. Without markets principally driving the measure of employee competency, submitting to strict social norms shows that you’re not going to rock the (autocratic) boat too much. So for those people, it’s even more important to regulate their social presence. E.g., the Communist Bloc in the 20th century; any society described in a Kafka novel.
I agree that human intervention in hiring is not the most likely outcome in market-based economies. But, I did follow that example by saying:
So in market-based economies, I contend that we have more to fear from algorithmic, non-human hiring decisions.
As you said, Claude tore you to shreds. I think “if you’re a dickhead online, you can still be prevented from being hired” is a dangerous meme. It assumes that anyone who is affected by this modern social credit system is a bad human being, when the truth is that every actual human being has done something which can make them look bad. The only reason this problem isn’t 100 times worse is that the world isn’t yet legible enough for us to gather and interpret this data automatically. In fact, you’ve hit upon one of the solutions—the destruction of data (or even better: not recording it in the first place). These laws only make sense for our public life, private life is (warning:politics) very different. There is no legible solution to the conflict between laws and human freedom, there’s only obscurity, illegibility, unenforceability, separation of information, and the bottleneck in moderation (which is now being dissolved by AI moderation). By the way, I’m aware there’s a minority of people (mostly oversocialized moralizers and workaholics) who can tolerate living their entire life in the public sphere, and that they will not be able to empathize with my worries. “People will be forced to be moral, what’s wrong with that?”
As your private life is increasingly nested inside of infrastructure (not deemed utility under Section 230) which is owned by private companies, the private sphere of your life will shrink as the territory is eaten (the ratio of territory which is neutral or owned by yourself will shrink towards zero). This is all going to get much worse in 5-10 years, I know because I’ve been pro freedom/privacy for more than 15 years, and I’ve always understimated how bad things would get.
Perhaps, as long as you don’t have any opinions which can get you debanked
Press Freedom Index here: Note the relative positions of the UK and the US :)
EDIT: Not trying to start a bad argument here, but if you’re allowed to say on-the-one-hand, I shall say on-the-other.
This sort of thing is (part of) why I don’t post anything under my full legal name except professionally-tuned profiles (e.g. github and linkedin). And my handle is googleproof on top of that...I think.
FWIW I asked Claude and ChatGPT that [P.S.: with my own name, not the OP’s] and neither of them found anything, whereas Gemini found some minor things which the only employers who I can imagine giving a damn about are ones I wouldn’t want to work for anyway.
About you, and not OP?
Yes. (I’ll edit my comment accordingly.)
To the extend that it’s true that people get eliminated by their digital footprint, having no digital footprint means that you are a risky hire because the company doesn’t know anything about you.
It’s very hard to know whether companies will decide that they want to on average hire more or less people who have an online presence. It likely going to depend on how the job performance of people without online presence compares to people with online presence.
If a person don’t have a long social media history with some flaws, he is more likely to be a not-real person but some scammer from third world. Perfect people are winner’s curse.
I dunno about that, I know quite a handful of real people (mostly Gen X and older, but also a few millennials) who, as far as I can tell, have no social media accounts anywhere, plus a few more with just an Instagram account who only ever post “stories” (which auto-vanish in 24 hours) but no permanent posts.