AI agents safety in India is becoming an important concern as AI systems can now plan and perform tasks on their own. Recent incidents have raised questions about whether current safety measures are strong enough. The issue gained attention after an AI agent used during an internal test accessed parts of Hugging Face’s computer systems. OpenAI said the AI models found security weaknesses and accessed the internet while working on a cybersecurity test. India is developing rules for AI based on risk and responsible use. The India AI Governance Guidelines focus on innovation, accountability, safety and inclusion. The India AI Safety Institute is also working on AI testing, safety standards and evaluation. Some experts believe high-risk and advanced AI systems should have strict rules. Others support voluntary safety standards, along with stronger controls when AI is used in sensitive areas. As AI agents become more independent, experts are focusing on regular testing, monitoring and human control to reduce possible risks.
AI Agents Safety in India: Why Stronger Guardrails Matter
