Bhathiya Barukanda
From utilising AI assistants like ChatGPT or Microsoft Copilot to tracking daily fitness routines with smartwatches, many individuals interact with AI systems or tools on a daily basis. While these technologies undoubtedly enhance convenience, they also raise significant concerns regarding data privacy.
Christopher Ramezan, an assistant professor of cybersecurity at West Virginia University, has studied how emerging technologies and various AI systems handle personal data, as well as how we can develop more secure and privacy-respecting systems for the future. He explains that generative AI software relies on vast amounts of training data to produce new content, such as text or images, whereas predictive AI uses data to forecast outcomes based on past behavior—like assessing your likelihood of achieving your daily step goal or recommending movies you might enjoy. Both types of AI can be employed to collect information about individuals.
How AI tools collect data
Ramezan explains that generative AI assistants like ChatGPT and Google Gemini gather all the information users input into the chat box. Every question, response, and prompt entered by users is recorded, stored, and analyzed to enhance the AI model. He notes that while OpenAI offers users the option to opt out of content usage for model training, it still collects and retains personal data. Although some companies claim to anonymise this data—meaning they store it without identifying the individuals who provided it—there remains a risk that the data could be reidentified.
Predictive AI
Ramezan explains that in addition to generative AI assistants, social media platforms such as Facebook, Instagram, and TikTok continuously collect data on their users to refine predictive AI models. Every post, photo, video, like, share, and comment—along with the time users spend engaging with each of these elements—serves as a data point used to create comprehensive digital profiles for every individual utilising the service.
Furthermore, Ramezan notes that smart devices, including home speakers, fitness trackers, and smartwatches, consistently gather information through biometric sensors, voice recognition, and location tracking.
Privacy rollbacks
The possibility of third-party access extends to smartwatches and fitness trackers, which track health metrics and user activity patterns. Ramezan notes that many privacy experts view the cloud storage of voice recordings as a method of data collection, particularly when it is used to enhance algorithms or create user profiles. This has significant implications for data privacy laws intended to safeguard online privacy.
Implications for data privacy
These developments raise significant privacy concerns for both individuals and governments regarding how AI tools collect, store, use, and transmit data. The primary issue at hand is transparency. Ramezan notes that many people are unaware of what data is being collected, how it is used, and who has access to it.
Companies often utilise complex privacy policies laden with technical jargon, making it challenging for individuals to fully comprehend the terms of service they agree to. Furthermore, people generally do not take the time to read these terms; one study found that individuals spent an average of only 73 seconds reading a terms of service document, which typically requires 29 to 32 minutes to read thoroughly.
Data gathered by AI tools may initially be stored with a trusted company, but it can easily be sold or shared with organizations that may not be as trustworthy. Ramezan points out that AI tools, as well as the companies operating them and those with access to the collected data, are vulnerable to cyberattacks and data breaches. These incidents may expose sensitive personal information and can be executed by cybercriminals motivated by financial gain, or by advanced persistent threats—often state-sponsored attackers who infiltrate networks and systems undetected, gathering information and personal data for potential disruption or harm.
Ramezan emphasizes that current laws are still evolving to address the challenges posed by AI and data privacy. For the time being, you should operate under the assumption that any AI-powered device or platform is collecting data on your inputs, behaviors, and patterns.
Utilising AI tools
While AI tools gather data on individuals, raising legitimate concerns about data privacy, they also offer significant benefits. AI-powered applications can enhance workflows, automate mundane tasks, and deliver valuable insights. However, it is essential to approach these tools with mindfulness and caution.
Ramezan advises that when using a generative AI platform that responds to the prompts you input, refrain from including any personally identifiable information, such as names, birth dates, Social Security numbers, or home addresses. In a professional environment, it’s critical to avoid sharing trade secrets or classified information. Generally, do not include anything in a prompt that you wouldn’t be comfortable disclosing publicly or seeing advertised. Once you hit enter on your prompt, control over that information is relinquished.
Ramezan also reminds us that devices that are powered on remain perpetually attentive, even in sleep mode. If you own smart home or embedded devices, turn them off whenever you require privacy for a conversation. A seemingly inactive device is still operational, constantly waiting for a wake word or signal. Unplugging a device or removing its batteries ensures that it is genuinely off.
Lastly, familiarise yourself with the terms of service and data collection policies associated with the devices and platforms you use. You may be surprised by what you’ve already consented to.
source:indiatvnews.com
