AI Dystopia Already Here: JP Morgan Chase Being Accused Of Using AI To Monitor Workers
Copy pasted if deleted
A warning for anyone working at or thinking about working at JPMorgan Chase & Co.
If you work at JPMorgan Chase & Co. or are thinking about working at JPMC, you need to know about their employee surveillance tool called WADU. WADU is an acronym for Workforce Activity Data Utility. Every employee at JPMC has a profile in the WADU database.
I think everyone expects their employer to track them to some extent. It is pretty standard practice for employers to monitor and run analysis on things like building badge swipes and the amount of time spent connected when working from home. It has also become very common place for employers to record audio and video at the office.
WADU is on a different level. It is an artificial intelligence & machine learning system for workforce human behavior. Starting at the moment you arrive to the building, WADU is tracking you using facial and speech recognition. Most JPMC offices and branches have been outfitted with some of the best HD AV security cameras.
Whenever you are at your desk, know that there is a HD camera tracking you the entire time. WADU uses the array of HD cameras at the office to monitor all of your non-verbal body language all throughout the day. The collected information is then fed into the AI/ML system and it is used to update your WADU profile in real time.
Every manager gets access to a dashboard that lists all the metrics about their subordinates. The productivity metrics about an employee start getting updated immediately after an employee logs into the system. If the employee is at the office, two bio-metrics are available, attention/focus and stress.
The bio-metric feeds are updated from the facial and behavioral tracking. Having a bad day? Stressed about something? WADU has already noticed this and alerted your manager. Can’t focus? Not working at your usual pace? WADU has already noticed this and alerted your manager. Did something you normally don’t do? It’s possible WADU flagged it as suspicious and alerted your manager.
WADU is also why they are pushing RTO or “return to office” so hard. Upper management does not care if some employees are more productive when they are working from home. They want everyone back in the office as much as possible so that their WADU profiles are being refined. Enhancing their insight into you is more important to them than better productivity from working from home.
A lot of teams are now required to come in two to three days per week. Director level and higher are required to come in four to five days per week. Upper management wants to see everyone at all levels back in the office five days a week. They have invested millions into the WADU system, and they want to get a return on that investment. That only happens whenever people are in the office as much as possible.
WADU is also watching and listening whenever you are working from home. If you installed Citrix Workplace on your own computer, and you permitted Citrix to access your web camera and microphone after login, you have connected those devices to WADU. If you are using an issued Chromebook, those permissions are already conveniently accepted for you.
You’ll notice that your web camera will flash right after login. This is not an “initial connection” flash. Your web camera just took a burst shot of pictures and sent them to WADU. The pictures will be scanned for anything deemed unprofessional or unsafe. Recreational drug paraphernalia, TVs, game consoles, and several other things are all flagged if detected in the pictures. If you see your web camera flash randomly, that was your manager or someone in security requesting a burst shot of pictures from your web camera.
You’ll also notice that your microphone will go hot shortly after login. Anything you say will be processed by WADU. All background noises will be processed by WADU. Say something bad about your boss or other superior? WADU flagged it. Say something bad about another co-worker? WADU flagged it. Have a moment of anger or frustration? WADU flagged it. These are just some examples, WADU is trained to detect a wide variety of keywords, phrases, and sound events. Your manager can also connect and listen to your audio feed live.
WADU is also able to detect keyboard pokers/bumpers and mouse jigglers/movers. It doesn’t matter if it is a completely external solution, WADU will be able to detect it by analyzing the repetitive input pattern. Your manager will be notified that your under suspicion of faking productivity. They will then connect to your session and see what is happening live. Action will be taken if the suspicion is confirmed.
WADU determines how productive you are by analyzing a variety of metrics about your session input. This includes words typed, mouse clicks, application activity, and many other things. The analysis also determines if someone is a unique contributor or if they are a regular worker. In overall rankings, unique contributors are always ranked higher than regular workers. The same analysis can also determine who is essentially dead weight. These people are ranked last.
You may have noticed at some point that you started getting job postings sent to your personal email. If you click on any of the links in these job list emails, your manager will get a notification on your WADU profile that you are actively looking for a new job. Even if it was just browsing, it can negatively affect the employee who clicked the link.
If you installed the JPMC workplace app on your phone, you have connected your phone to WADU. The workplace mobile app will collect a variety of information from your phone and use it to update and refine your WADU profile. Right now, the only way to reserve a desk at the office is to use the workplace app. The web version of the desk reservation system is still “coming soon” and you are pushed to install the app on your phone. It will probably still be “coming soon” in 2040.
Upper management pushes a narrative that all this surveillance is required to safeguard the firm against insider threats. While that may be partially true, the main reason is to train and refine the AI/ML system. They want every employee profile to be as accurate and as detailed as possible.
They say we are not supposed to use anything from an employee’s WADU profile to make employment decisions. It is kind of hard to ignore a ranked list of subordinates with productivity forecasting.
Here is how you can minimize the privacy invasion while working at JPMC:
- Keep your web camera covered and microphone muted unless on a Zoom
- Most modern laptops have multiple web cameras and microphones, make sure you are aware of all potential recording devices on your computer
- Use your own desktop or laptop at home instead of an issued Chromebook
- Do not grant Citrix Workplace access to your web camera or microphone
- Use Zoom outside of Citrix when you are at home
- Consider getting a phone and number solely for work purposes
- Do not install the JPMC workplace app on your personal phone
- If you need the workplace app, factory reset an old phone and install it on that
- Know there are no blind spots, you are always on camera at the office
- Know there is no privacy, all conversations are recorded everywhere at the office
- Know all of your behavior at the office affects your WADU profile and ranking
- It would be a good idea for you to print a copy of all performance reviews at home
I know this is a Reddit post, and it might all be a hoax but its not too far off from what Amazon has already been doing to their warehouse and distribution workers.
Amazon’s surveillance software gathers data from worker scanners which are used to scan an item every time it moves from a different part of the logistical process. This means that surveillance is integrated at every level of the workplace, and performance is being monitored on this basis. European Amazon boss Brian Palmer admitted to parliament in December that workers can be fired for not meeting targets, which include the requirement to hit a rate per hour, and to not go over a certain idle time, also known as Time Off Task.
With productivity targets generated by an algorithm, workers in Amazon warehouses are under constant scrutiny and pressure to work faster and harder. Garfield Hylton is a 59-year-old Universal Receive – which means he processes and checks products for damage before they’re taken to be dispatched – who has been working at the Coventry warehouse for almost five years, and is currently training to be a GMB rep.
Featured on Barstool
The cost-benefit for a bank to invest a couple of million into software to monitor their workers who have all started to work mostly out of the office sounds like a no-brainer from a management perspective, but it is going to hit a whole new level of 1984 corporate surveillance if the government does not step in. The government is technically not allowed to take such oppressive action, but large multinational corporations may end up being the real entity to usher in a 1984-type scenario. Microsoft Teams and slack have always been monitored by employers, so this type of scanning of every email, message, etc has already been compromised but monitoring employee's ability to try to search for a new job while currently employed is an unprecedented advantage in the job market. From the sounds of the ability of this new program, it literally will ensure that you cannot get away with fucking around at work. Hopefully, the robot realizes that you cannot hold humans to such a high standard; everyone is going to fuck around to a certain degree. Even a completely studious robot-like nerd (referring to humans) has to go take a coffee shit every now and then. Hell, even the Wall Street bros who run on Adderall and cocaine are going to get pinged by this. This program is going to convince employers that robots can do all these jobs way faster and more efficiently then these resource-sucking carbon beings that are inefficient due to "Bodily functions". Andrew Yang tried to warn us this was coming, he should have listened. Hey, I am an anti-AI blogger that AI probably can't replace. There is a reason the term "Yuppie" doesn't get thrown around that much anymore, it's because a lot of union manual laborers are making just as much if not more than a lot of these "Yuppies" and they don't give AF that they are about to lose all their jobs to AI while still having student loans while they completed trade school and get benefits.