CERT-In ne AI agents ke liye security guidelines release kar di hain. Ab AI ko 4 alag-alag risk tiers mein baanta gaya hai aur 'Human-in-the-loop' zaroori hoga.
AI par Govt ki kadi nazar
CERT-In ne ab AI systems ke liye ek formal security framework laa diya hai. Iska main maqsad hai AI agents ke wajah se hone wale data leaks aur unauthorized access ko rokna. Ab se AI systems ko 4 risk tiers mein divide kiya jayega, jismein 'Critical' systems ke liye sabse strict rules honge.
Kya badal raha hai?
Sabse bada change hai 'Human-in-the-loop' mandate. Iska matlab hai ki ab koi bhi bada decision AI apne aap nahi le payega; uske liye human approval zaroori hoga. Iske saath hi, companies ko ab 180-day ke logs maintain karne honge aur kisi bhi incident ki reporting 6 ghante ke andar karni hogi.
Investors aur Companies par asar
Jo companies AI infrastructure mein heavy investment kar rahi hain, unke liye compliance cost badhna tay hai. Cybersecurity aur legal compliance ke liye ab zyada paisa kharch karna padega. Halanki ye guidelines abhi non-binding hain, lekin market experts ise future mein aane wale sakht rules ka trailer maan rahe hain.
Ab dekhna ye hai ki IT companies apne current architectures ko kitni jaldi in standards ke hisaab se update karti hain.