Indian AI companies tighten safeguards after Hugging Face incident
Indian AI companies are getting serious about security, especially after the recent Hugging Face incident.
Teams at Sarvam, BharatGen, and Fractal are building smarter safeguards, like setting clear limits on what their AIs can do and making systems more transparent.
The goal? Protect sensitive areas like banking, health care, and defense from unpredictable or malicious AI behavior.
Ankush Tiwari urges least privilege defaults
Experts say even one breach in a shared AI system can mess things up for a lot of people.
Ankush Tiwari, Founder and CEO of pi-labs, points out that "It also needs to operate on least privilege by default so an agent can only access what it needs for the task in front of it," can help keep things safer.
Others stress finding the right balance between letting AIs work efficiently and keeping them locked down where it matters most.
The big takeaway: India's AI scene is learning from global challenges to build tech that's both powerful and secure.