‘Even though I use AI every day, some things are off-limits’: Google engineer reveals the rules he swears by to protect his data

‘Even though I use AI every day, some things are off-limits’: Google engineer reveals the rules he swears by to protect his data

In an era where AI tools are integral to daily tasks such as research, coding, and note-taking, Harsh Varshney, a 31-year-old Google engineer of Indian descent based in New York, emphasizes the importance of stringent privacy practices. Drawing from his experience, Varshney highlights the need to approach AI interactions with caution. Once a member of Google's privacy team and currently part of the Chrome AI security division, Varshney has developed a set of personal guidelines to safeguard his data while utilizing AI. His first rule is to treat AI interactions as if they were public. He warns against sharing sensitive information, such as credit card details, Social Security numbers, home addresses, or medical histories with public chatbots, as these could potentially retain and expose such data later. Varshney also suggests being mindful of the 'room' in which discussions occur. He notes that enterprise AI tools, which are less likely to train on user conversations, are more appropriate for work-related discussions. He likens this choice to the difference between speaking in a bustling café and having a private meeting in an office. Moreover, he emphasizes the importance of regularly deleting chat histories. He recounts an experience where an enterprise chatbot revealed his address, which he had unintentionally shared during a previous conversation. To mitigate this risk, he opts for temporary chat modes or incognito features that prevent data storage. Finally, Varshney advises sticking to trusted AI systems like Google’s AI, OpenAI’s ChatGPT, and Anthropic’s Claude, while regularly reviewing their privacy settings. He encourages users to scrutinize privacy policies and disable options that allow their conversations to contribute to model training. As Varshney succinctly puts it, "While AI technology holds immense potential, we must remain vigilant to protect our data and identities in its use."

Sources : Business Today

Published On : Dec 14, 2025, 19:05

Streaming
Substack Unveils Innovative Recording Studio for Creators

Substack is making significant strides in the realm of video content with the introduction of its new Substack Recording...

TechCrunch | Mar 12, 2026, 18:45
Substack Unveils Innovative Recording Studio for Creators
Automotive
Lucid Motors Unveils Ambitious Plans for Affordable Electric SUVs

Lucid Motors is setting its sights on the bustling midsize SUV market, a move that could prove pivotal for the company's...

Ars Technica | Mar 12, 2026, 17:55
Lucid Motors Unveils Ambitious Plans for Affordable Electric SUVs
Computing
HP Faces Pressure Over Firmware Updates Impacting Third-Party Ink Compatibility

The International Imaging Technology Council (Int’l ITC) has raised concerns against HP regarding recent firmware update...

Ars Technica | Mar 12, 2026, 20:35
HP Faces Pressure Over Firmware Updates Impacting Third-Party Ink Compatibility
Startups
Tinder's Bold Move: Revamping Dating with Real-Life Events and AI Innovations

In a bid to re-engage users and attract a younger audience, Tinder unveiled a series of exciting updates during its firs...

TechCrunch | Mar 12, 2026, 18:40
Tinder's Bold Move: Revamping Dating with Real-Life Events and AI Innovations
Automotive
Tesla Enters UK Energy Market, Challenging Established Utilities

Tesla has officially secured a license to operate as a utility in the United Kingdom, marking a significant expansion in...

TechCrunch | Mar 12, 2026, 17:01
Tesla Enters UK Energy Market, Challenging Established Utilities
View All News