AI Fought the Law, and the Law Won
What happens when both the right and the wrong sides of the law have access to the same transformational technologies? Find out This Week in Fraud Trends, July 5, 2019.
Social media has a tendency to symbolize everything both wonderful and terrible about the digital age. The freedom, the connectivity, the connectedness … the crime.
Everything that makes social media great is also what makes it vulnerable, and no one knows this better than the fraudsters who continually exploit its weak spots.
While data breaches may often get pride-of-place when it comes to the headlines, we learned from Bloomberg this week of a different malicious exploitation of a social media platform:
“In one of the largest malware campaigns to exploit Facebook Inc., a suspected Libyan hacker lured tens of thousands of people into exposing personal information and granting access to personal devices, Israeli cyber security company Check Point Software Technologies Ltd. said.”
While Facebook often seems to find itself at the center of fraud controversies, it’s important to remember they’re also working hard to solve some pressing challenges. We read a Wired article this week that showed Facebook adopting one of our own recommended tenets for comprehensive fraud protection — contextual detection.
“Facebook is now testing a new content moderation workflow that prioritizes this context-first approach to the review process. Whereas previously moderators made a decision about whether or not to remove a post and then answered a series of questions as to why, the pilot program for its US hate speech enforcement team reverses the order of that process: When assessing whether a post has broken the rules, reviewers are asked a series of questions first, then prompted to make a decision.”
Contextual detection is one of the key capabilities that makes proactive fraud prevention possible — something we wrote about recently on our blog …
“Unsupervised Machine Learning (UML) offers fraud teams a new superpower — the ability to deploy highly accurate detection models without the requirement of historical data or preexisting labels. This approach can feed into an overall strategy that prioritizes holistic analysis and contextual detection.” — Yinglian Xie, CEO, Co-Founder, DataVisor
… and right here on Medium:
“Account-level detection, on the other hand, focuses on the accounts themselves — how and when they’re created, and, most importantly, the context in which they’re created. Using a solution like DataVisor’s dCube, you can detect fraudulent activity at the point of account creation.”
While social platforms are indeed a favorite fraudster target, cloud services are being abused with alarming frequency, as we learned this week from TechTarget:
“According to the report, which tracked 5,334 new phishing kits deployed to the web so far this year, the tactic of hosting phishing domains on public cloud services has ‘grown significantly’ this year.”
As we think through vulnerabilities like those exposed above, it’s important to remember that fraud actors have access to all the same technologies we do, which means that when we continue to rely on outdated systems and tools, we automatically give them the upper hand. Dark Reading raised alarms on this topic earlier this week:
DataVisor CEO Yinglian Xie, in a PYMNTS podcast episode published this morning, addressed this challenging dilemma head on:
This is exactly the point we raised earlier in this article; AI is truly a transformational technology, but it’s not exclusive to the right side of the law — fraudsters can use it too.
To learn more about how AI can be used to stay AHEAD of the fraudsters — particularly as it pertains to social platforms — we recommend Defeating Mass Registration with Unsupervised Machine Learning by Sean McDermott.
We also recommend you check in next week for another episode of This Week in Fraud Trends!
