Home CompanyCommunity Spokeo Announces 2023 AI Scholarship Winner
Home CompanyCommunity Spokeo Announces 2023 AI Scholarship Winner

Spokeo Announces 2023 AI Scholarship Winner

by Spokeo

Spokeo, a leading people search engine, proudly reveals the winner of the Spokeo Artificial Intelligence (AI) Scholarship—a groundbreaking initiative supporting exceptional students intrigued by AI’s transformative potential. From a rigorous selection, Adrianna Pinzariu of DePaul University Jarvis College of Computing and Digital Media‘s class of 2025 emerged as the deserving $5,000 scholarship recipient.

Adrianna’s essay highlighted her leadership via SecureAI, an AI tool aiding law enforcement against human trafficking. SecureAI analyzes hotel websites to enhance trafficking investigations. Her Flock campus safety app, another AI creation, advances data collection and detection. Remarkably, Flock won $6,000 in the 2023 DePaul Purpose Pitch competition.

Applicants could use AI in essays with specified prompts. Spokeo creatively employed ChatGPT for initial assessment of the essays. However, Judges selected finalists and the winner.

Spokeo logo

Who's Calling Me?

Search any phone number to learn more about the owner!

Check out the video below where CEO Harrison Tang sits down with the talented to discuss her background and interest in AI before notifying Adrianna that she’s the winner.

Building AI for a Safer Tomorrow: AI’s Role in Combatting Human Trafficking

By: Adrianna Pinzariu

AI Technology is not merely a tool, but a transformative motivator with the potential to be used to reshape our world for the better. With its current applications in combating societal issues, aiding in sustainability efforts, or automating daily tasks, it’s evident that we’ve only just harnessed a small percentage of its true capabilities. This untapped potential is emblematic of why I took up coding in the first place; my dream of harnessing this technology to establish a nonprofit dedicated to combating human trafficking. My journey in this field began last year when I was delivering award-winning speeches on the use of AI in combating human trafficking and the issue of algorithmic bias. Although my experience with coding at the time was minimal, I’ve always been driven by the concept of social purpose and found myself investing all my time in devouring research papers and learning to code. I’ve looked into how Tools like Traffic Jam by Marinus Analytics and Spotlight by Thorn leverage machine learning and facial recognition to identify victims in online advertisements, drastically reducing investigation times. Meanwhile, Microsoft’s PhotoDNA service uses hash matching to detect and report known illegal images related to child exploitation, helping to eliminate such disturbing content from the digital world. And on top of that, these tools are simultaneously utilizing pattern identification to identify and take down transnational trafficking networks.

However, despite these advancements, there’s still a lot of room for improvement due to the ever-advancing nature of the A.I. industry. This realization caused me to recently lead the development of SecureAI, an AI-driven human trafficking indicator tool for law enforcement agencies. The tool sifts through large amounts of data from hotel websites, which are typically notorious hotspots for trafficking activities, and picks out pieces of information that when pieced together could hint towards an ongoing trafficking operation. SecureAI then compiles this information and generates a detailed report which can be readily utilized by law enforcement agencies for further investigation and action. Ideally, the key was to extend the capabilities of SecureAI and integrate it with a venture I have a personal connection with, Flock. What initially started off as the overall winning hackathon project for University of Illinois Urbana Champaign’s WECE Hacks Hackathon, had transformed into a startup once my team and I had recognized that the project had potential to change people’s lives. Flock, a women’s campus safety app, is a startup that’s goal is to eradicate the 1.4 million sexual assaults that happen yearly at college campuses. Powered by a tool to send users on the AI optimized safest paths based on crime data, profile-based matching system, and emergency alert system, Flock’s mission is to envision the security of college students, day and night.

While I envision SecureAI transitioning into its own venture, garnering more dedicated researchers and expanding our knowledge base over time, one of the biggest constraints of combatting the complex, multifaceted issue of trafficking is the scarce amount of data available. The accuracy of AI algorithms and quantity of data share a linear relationship, as in the more data you input the more accurate your AI model becomes. While we all have multiple groups that want to combat trafficking, the process is hampered by a lack of a central entity to unite all of the data. This is when the integration of Flock is crucial, as a campus safety app, it gathers invaluable real-time data on user routes, reported safety incidents, and user feedback. This collaboration is critical to fill in the void of a necessary centralized data collection system, thus enhancing the SecureAI algorithm’s trafficking detection capabilities and overall contributing to the global fight against this evil.

Moreover, as AI such as SecureAI become more publicly available, it’s critical to address the downsides associated with this early on, along with potential ways to mitigate these risks in the future in order to ensure this technology protects the people it’s designed to serve. To begin, there are major data and privacy concerns such as its potential misuse or data breaches. One way of circumventing this is already implemented in SecureAI, through our anonymization algorithm. The way this works is that an image is inputted of personally identifiable information such as faces or license plates. To get rid of this personal information, we apply a masking algorithm with object detection to identify what item to remove, and then apply an AI inpainting model to remove the item and fill it in with a background to ensure privacy. This makes it appear as if the personal details were never there to begin with. However, dealing with potential misuse extends beyond the anonymization algorithm. For instance, we’ll encrypt the original images and establish strong security protocols when it comes to data storage and access. For instance, we’ll establish a data retention policy with corresponding law enforcement agencies to ensure the data is only kept as long as needed and then permanently deleted.

Another risk would be the use of these technologies getting in the hands of the traffickers themselves. Thus, the next steps would be to work closely with law enforcement agencies when coming up with strategies to mitigate this. One proactive measure currently being implemented in modern machine-learning technologies is employing predictive analytics. By monitoring suspicious patterns to identify trafficking networks early on, this provides law enforcement agencies lead time to prepare and take down these operations. This new data then gets fed back into the model training it to improve its accuracy, allowing law enforcement to stay a step ahead of traffickers. Furthermore, to ensure this AI technology doesn’t get into the wrong hands, there will be strict access controls maintained where only authorized individuals have access to the technology. To enhance the security of the data, I’ve taken the initiative to build up my skill set through programs such as the selective Major League Hacking (MLH) Fellowship, where I was chosen to participate in the Site Reliability Engineering Track. Through this program I’ve been able to learn different advanced cybersecurity techniques such as the application of monitoring systems to identify potential data breaches early on while also implementing secure system design to improve the security of both SecureAI and Flock.

As we delve deeper into an era where AI’s capabilities are expanding at an exponential rate, it is crucial to adapt and match pace with the complexity of the problems we’re working on solving. For me, this means building the most up-to-date AI Algorithms to help advance campus safety and combat human trafficking. However, my vision is just a small piece of what will make up the future once we harness AI’s potential to its fullest. So I ask, are you ready to join us in shaping this new era?

AI Application Used:

  1. ChatGPT

Prompts Provided:

  1. How is AI currently being used for socially good purposes?
  2. How does predictive analytics reveal transnational human trafficking networks?
  3. What are some potential security vulnerabilities in developing an AI Algorithm to combat trafficking?
  4. What are the risks of an AI human trafficking indicator tool?
  5. How can we keep AI from falling into the wrong hands when combating human trafficking?
  6. How will the use of building a central entity to collect data to combat trafficking enhance anti-human trafficking efforts?
  7. What are some good data retention policy ideas?
  8. Why is human trafficking such a difficult issue to solve?
  9. What are common data breaches for tech companies?
  10. How can we keep the AI human trafficking indicator tool, SecureAI, from potential privacy concerns or data breaches?