The Google Engineer’s 4 Rules for Staying Safe While using AI
AI has quickly become a silent partner in our everyday lives, and I can no longer imagine life without AI tools.Every day they help me with extensive research, note-taking, programming and online searches.
But my job means that I am very aware of the privacy concerns when using AI. I have been working at Google as 2023 and spent two years as a software engineer in the data protection team, where I built infrastructure too protect user data. I then joined the Chrome AI security team, where I help protect Google Chrome from malicious threats, such as hackers and those using AI agents for phishing campaigns.
AI models use data to generate useful answers,and we as users must protect our private facts so that malicious actors such as cybercriminals and data traffickers cannot access it.
Hear are four habits I’ve developed that I believe are essential to protecting my data when using AI.
Read too: Ex-hacker warns of 3 new cyber threats
Why You Shouldn’t Use ChatGPT at Work
ChatGPT and other large language models (LLMs) are incredibly useful tools, but using the free, public versions at work can be risky. While they can help with tasks like drafting emails or summarizing documents, you need to be aware of the potential downsides, especially regarding data security and intellectual property.
The biggest concern is data privacy. When you input information into a public chatbot like ChatGPT, that data is used to train the model. This means your company’s confidential information could inadvertently become part of the public knowledge base. Think of it like discussing sensitive topics in a crowded coffee shop where you could be overheard, as opposed to a confidential meeting in your office that stays in the room.
There have reportedly been cases where employees have accidentally shared company information with ChatGPT. If you’re working on unpublished company projects or trying to get a patent, you probably don’t want to discuss your plans with a non-enterprise-grade chatbot because of the risk of a leak.
I don’t discuss projects I’m working on at Google with public chatbots. Instead, I use an enterprise model, even for tasks as small as processing a work email. I feel much more comfortable sharing my information because my conversations aren’t used for training, but I still minimize the personal information I share.
Read too
