Workplace spyware when everyone works from home

You may not have noticed (yet) but WFH has changed forever. Employee surveillance tools that were once used in relatively narrow ways and focused on specific industries (such as to detect insider trading in the finance industry or for time tracking in consulting businesses) are increasingly being adopted by companies that have sent employees home to work during the pandemic crisis.

Modern AI-powered employee monitoring tools are extremely powerful. There are many options and the tools are in demand; some vendors have seen inbounds triple since WFH became mainstream, according to a recent Bloomberg article. Common features of surveillance tools include:

• screen shot capture and monitoring, including automated monitoring of sites deemed to be “not work relevant” or generally non-productive.
• automated monitoring of chats and emails for sentiment or “unsafe” content.
• productivity tracking and comparison tools.
• automated check-ins and nudging for when online activity is low or “off-target.”
• behavioral analysis based on machine-readable activity.
• classifying all online behavior into productive versus unproductive.
• monitoring employee via their computer camera.
• high risk activities such as printing behavior — e.g., printing a confidential client list followed by a resume, which may indicate a flight risk.

Part of the deal that comes with being an employee is being monitored. But monitoring a newly remote workforce with tools that were deemed unnecessary when everyone was in the office is a significant shift. Just a few weeks ago, chatting at the water cooler was likely to be viewed as “building rapport” or “cultivating influence.” Contrast this to now, where WFH surveillance might deem any time away from a screen as “unproductive.” AI is classifying behavior and employers risk losing the nuance of what makes for high performance and, ultimately, employees’ trust.

The tradeoff is framed as trust versus prudence — for bosses worried about subordinates slacking off, perhaps taking an extended spring break, some claim it is imprudent not to use these tools. Surely, being able to measure people’s behavior in a mathematical and tangible way is the only way to manage productivity in a remote workforce? Aren’t these tools also useful to the employee, who can often (not always) customize the tool and use the output to improve themselves? That’s the marketing spin anyhow.

An alternative view is that the use of these tools is a measure of company maturity, and a low one at that. This is according to one of the gurus of distributed workforces, Matt Mullenweg, founder of Wordpress and Automattic. Mullenweg spoke recently on Sam Harris’ podcast, where he talked about the 5 levels of distributed teams and the future of work.

The five levels of distributed teams are:

1. non-deliberate action — employees do a bit at home but wait until they’re back in the office to do anything substantive.
2. recreating office online — people expected to be at work 9-5, workplace “spyware” and lots of big video meetings and long email info-shares. Most companies today are here, according to Mullenweg.
3. adaptation — companies invest in making people more productive: better desk setups (lighting and audio) because little things matter, people get used to real-time collaboration in shared docs, meetings are for decisions, personal relationship building and check-ins not broad-scale communication.
4. asynchronous communication — communication follows urgency, employees allowed to reach state of “flow” themselves and in their own way. Presence and productivity are decoupled and no longer a measure of performance.
5. “nirvana” — distributed team works better than any in-person team.

The big question everyone wants the answer to is whether WFH will become a permanent thing or whether people will head back to the office once this is all over. Once you look at this through the lens of the 5 levels, it’s pretty clear that staying stuck in level 2 is a non-starter. Either people will head back to the office or they’ll move to level 3. Staying subject to “spyware” with none of the benefits of personal relationships in a co-located physical environment just won’t be tenable.

So how should employers use this time and the AI tools available? Yes, there are good intentions behind these tools — some people really appreciate being able to learn about their work habits and to be nudged to get off Facebook. Some sites, such as Sneek, started out as a product that provides “human contact for remote teams.” The company’s “wall of faces” product “lets you see your team mates all day and start instant video chats with a single click.” Of course there are surveillance-heavy ways to use this; if a coworker clicks on someone’s face, Sneek's default settings will instantly connect the two workers in a live video call, even if the recipient hasn't clicked "accept."

Welcome to micromanagement hell. Chances are, none of these employees were subjected to supervisory visits to their desks every few minutes while at work. There's no reason for them to be subjected to it now just because they're working from home. The only reason this is happening is because it can happen. It's seamless, automated, and makes zero physical or mental demands from their employers. - Tim Cushing, Techdirt

Organizational psychologist and expert on workplace behavior and culture Adam Grant should repost (he’s much more famous now) a story he wrote in Psychology Today back 2013 called “Instead of Monitoring Employees, Try Motivating Them.” While the article was primarily focused on workplace theft, there are vital insights that are 100% applicable to modern workplace surveillance:

• surveillance sends a clear message to those under surveillance that they are not trusted. This breeds resentment, especially among high performing employees.
• the act of surveillance makes people more suspicious. If people look for bad behaviors, they will be sure to find some. Confirmation bias kicks in and pretty soon, perceptions will be dominated by untrustworthy employees. People become skeptical of others’ motivations and the whole system of trust can collapse. For remote workers, this is especially problematic as people start to easily mis-interpret the intent behind wording in an email or slack message, seeing Mal-intent where their is none.
• the decision to implement a surveillance system communicates to employees that what you are trying to detect (eg low productivity) is common. This can be self-fulfilling and people may become less committed to internal standards of integrity in the workplace.

AI is different. WFH surveillance is not so much a prudent choice as a passive, lazy one. The challenge isn’t going to be “when will we go back to work?” as much as it is “what will the new workplace be?” COVID is terrifying and catastrophic. Employers that recognize that even their best, highest paid knowledge workers won’t get through this unscathed, who set the scene for a more flexible, responsive, autonomy-respecting - and frankly, more human, workplace will ultimately be the employers of choice.

Also this week

• Excellent podcast from Recode Decode with Adam Grant on the future of work.
• Answering key questions on use of Apple/Google COVID tracking app from The Verge.
• Clearview AI has strong ties to the far-right and prominent neo-Nazis, according to Huffpost.
• ‘Unconscionable’: Landlord service Naborly draws criticism for creating ‘blacklist’ of tenants who missed April 1 rent - the company will retrain its tenant scoring models to stop people from renting again. This is blacklisting at scale.
• Selection of vendors of workplace surveillance tools FYI: Interguard, Time Doctor, Teramind, Vericlock, Inner Activ, Activ Track, Hubstaff, Sneek.