PayPal's AI private eye

Sign up to uncover the latest in emerging technology.

Plus: Intel’s automated manager; Apple and Google’s homestyle metaverse

Happy Monday and welcome to Patent Drop! 

Today, we’re diving into PayPal’s idea to sic AI on fraudsters, Intel’s tech to become your entire company’s automated planner, and Apple and Google’s different plans to get to know the inside of your house.

But first, a word from our sponsor, the staffing pros at Oceans. They’re in the business of helping you grow your team (and your business) with experienced global talent at 50-80% less than US-based candidates. Marketing, sales, finance… Oceans’ crème de la crème talent have 5-10 years working experience with global companies like Deloitte, PWC, Sysco, and more. Schedule a call with an Oceans expert today and grow your team for less. 

Alrighty folks, let’s take a look. 

#1. PayPal’s hack spotter

PayPal takes fraud so seriously that it’s turning AI into Sherlock Holmes. 

The company filed a patent application for tech that detects fraud using machine learning. According to the filing, this system will rely on an AI “designed and trained to recognize the patterns in fraudulent attempts” to decide whether or not to allow or deny a user’s request, such as access to an account or making a payment. The system is trained on previous user activity, taking into account things like who they commonly make transactions with, browsers or devices used, and times of access. 

For example, if a PayPal account for a person based in San Francisco suddenly makes a payment to a user in London at 4 a.m. PST, despite never having sent funds to anyone outside of the U.S. previously, the AI would flag that as suspicious, making the user verify in some way that the transaction was actually made by them.  

“Security is a universal problem in computer systems,” PayPal said in its file. “Malicious actors may, for example, fraudulently use a legitimate user’s compromised account to access the computer system and engage in transactions. A compromised account may be used to access secure electronic resources, transfer money, or make purchases.”

PayPal’s patent seeks to limit payments fraud and account takeover attacks; two common issues among fintechs, especially peer-to-peer payment apps. According to the Digital Trust and Safety Index report by fraud detection platform Sift, account takeover fraud increased 71% for fintech accounts between 2021 and 2022, and 79% in crypto-related accounts. 

But AI can go a long way in making platforms safer, said Mary Writz, SVP of product at Sift. An AI-based algorithm can take in tons of information about a customer’s journey, such as time of login, location, and payments made, and more quickly detect whether or not certain activity is fraud. And of course, because it’s AI, fraud detection will only get more and more precise over time.

For a company like PayPal, implementing AI for fraud is a perfect fit. Given that it handles tens of billions of transactions a year, using AI is the best way to quickly and accurately detect fraud among the thousands of payments it sees every minute. Plus, it has plenty of data to train its AI on, said Writz, meaning it won’t take long for its system to get good at tracking down fraudsters. 

“The only way you can operate at scale and latency needed is to move a lot of the advanced 

algorithms to AI and machine learning,” said Writz. “More modern providers are moving to AI-driven approaches in cybersecurity and in online fraud detection.” 

And if an incident does slip through the cracks, PayPal’s got an AI for that, too: The company wants to patent a deep learning model that can label and handle customer complaints. The future truly is now. 

#2. Intel’s planning multitool

Intel might make your manager obsolete… you’re welcome.

The company is seeking to patent tech for the “management of collaborative teams,” The system, which Intel called a “managing module,” is essentially an all-in-one tool for managing organizations and workflow.  

In the filing, Intel laid out several functions of this module: It can automatically assign team members to a project by reviewing documents they’ve collaborated on and any teams each individual member may be associated with; adjust permissions to documents based on those team members; assigning team members to tasks based on their skills and expertise; and make changes to documents based on messages to one another. 

“Collaborative teams may create a logistical challenge for each worker on the team,” Intel said in its application. “It may be difficult for a worker to keep track of the various teams of which the worker may be a member … Additionally, it may be difficult to ensure that each worker is working from the most up-to-date content.” 

One thing to note: Intel mentioned that the module could integrate with third-party applications, specifically bringing up the Microsoft Office suite as an example. 

Intel’s tech is just another example of how automation is making its way into the workplace. As we talked about in Patent Drop last week, we already see automation starting to take hold of the HR and recruiting process. And while workflow and document management tools aren’t anything novel, tech like what Intel lays out takes it a step further than what’s on the market today. 

Intel’s tech also solves a problem that has long plagued workplaces: Getting the right people on the right tasks, said Drew Dillon, product leadership consultant and CEO of community automation startup Burb. Finding the people on your team with the right expertise to take on certain projects and getting them the most up-to-date information to do so can often prove to be a frustrating task, he said. 

The problem spans industries and organizations of all sizes, Dillon added, from startups to the U.S. government. He gave the example of the transfer of classified information in the military: “Despite the actual classification of the information, how does one, if they’re a troop on the ground in Afghanistan, get a piece of intel back to a beat cop in New York if it’s relevant to helping prevent an incident? it’s a really old challenge that Intel is getting at here.”   

While you might wonder why Intel is researching it, the company has had partnerships with Microsoft practically since the dawn of tech, ranging from cloud services to DARPA contracts. Though the companies’ famous “Wintel” alliance on personal computing products frayed more than five years back, their interests are seemingly still aligned – at least in some areas. 

So keep an eye out: a robot task manager might show up in your Teams chat sooner or later. 


Expand Your Headcount, Shrink Your Payroll

When we first jumped into the newsletter game, we ran lean to keep costs down. But soon we learned that until we expanded our team, we would never grow. 

If only we knew about Oceans back then. 

Used by founders and corporations alike, Oceans surveys your needs and staffs you with top-shelf executive assistants, sales, and finance professionals — all for 50-80% cheaper than U.S.-based candidates. 

Oceans maintains a robust pool of seasoned talent:

  • Backgrounds in global companies like HSBC, Deloitte, PWC, and The World Bank

  • Candidates carry 5-10 years experience in various operational roles

  • Marketing, sales, accounting, you name it

To put simply, you could onboard an Executive Assistant with 7+ years experience working at The World Bank in ops and marketing for roughly $3,000/month. 

Schedule a call with Oceans today and join the likes of White Star Capital and Philadelphia Eagles’ Ndamukong Suh. Dive in here. 

#3. Your (virtual) home away from home

Both Google and Apple filed patent applications for tech that learns the layout of your home so you’re not falling over your couch with that VR headset on. 

Let’s start with Google. The company is seeking to patent tech for creating “three-dimensional maps of an indoor space.” This tech works by using depth sensors on your mobile device’s camera to “scan” the desired areas in your home and “generate realistic virtual rooms, which a user can interact with in an immersive and realistic experience.” Google also mentioned that this tech has AR applications, such as helping it more easily place virtual objects in a real-world environment.

“Scanning the indoor space successfully has multiple requirements and sometimes these requirements are beyond the control of a user,” Google said in its filing. “If all of the requirements are not met, a poor scan may result, and the corresponding 3D map may be incomplete and/or distorted. This is especially true when scanning a large indoor space with multiple areas.”

And in true Big Tech fashion, Apple is working on practically the same thing. The company filed a patent for “floor plan generation based on room scanning.”  Similar to Google’s patent, this system would take in-depth sensor data captured by a user walking around the room and scanning their surroundings to create a 3D replica of your home that can be transposed into an extended reality environment, like VR or AR.