OpenAI’s new model for defensive security; NVIDIA’s models for quantum computing research

Hello,

India is working actively towards sovereign AI to reduce reliance on external technology. 

As part of this push, India’s AI poster child Sarvam AI and IITM Pravartak Technologies Foundation are collaborating for sovereign AI deployment. The collaboration strives to deliver production-ready AI applications aligned with India’s data and governance needs, with a focus on multilingual models and local infrastructure.

Still on AI, India is among the world’s most advanced AI users but adoption is uneven and highly concentrated in a handful of cities, says an OpenAI report. 

In other news, One 97 Communications Ltd, which operates the Paytm brand, has become a majority Indian-owned and controlled company after domestic investors increased their stake to 50.3% as of March-end 2026.

This marks an important structural change in ownership for the fintech firm, with domestic shareholding rising steadily in recent quarters, reflecting growing investor confidence.

Meanwhile, after a failed bid for Warner Bros. Discovery, Netflix plans to focus on ads and content. Investors expect the streaming giant to now refocus on sports and other live events to boost ad revenue. Netflix will report its first quarter earnings on Thursday, the first results since it walked away from the deal. 

Lastly, lawyers in the US are warning clients not to treat chatbots like their trusted confidants as their chats could be used against them. 

In today’s newsletter, we will talk about 

• OpenAI’s new model for defensive security 
• NVIDIA’s models for quantum computing research

Here’s your trivia for today: Which company was originally incorporated as Cadabra Inc.?

Insight 

OpenAI’s new model for defensive security 

ChatGPT maker OpenAI has announced GPT-5.4-Cyber, a specialised version of its latest AI model that’s designed for defensive security work. The development comes a week after rival AI firm Anthropic unveiled its own security model, Claude Mythos Preview.

The introduction of GPT-5.4-Cyber highlights what OpenAI calls the next era of cyber defence. Unlike general purpose models that are often trained to refuse any requests involving computer code that looks even slightly suspicious, this new variant is described as being cyber-permissive.

Key takeaways:

• The AI has lower refusal boundaries for legitimate security tasks. It is designed to understand that a security professional trying to find a flaw is doing something helpful, not harmful.
• A notable feature of this model is its ability to perform binary reverse engineering, a technical process wherein an expert looks at compiled software, the version of a programme that a computer runs, to understand how it works without having access to the original blueprints or source code. 
• The AI can help defenders spot hidden malware or vulnerabilities in software that was previously very difficult to analyse.

Technology

NVIDIA’s models for quantum computing research

Tech giant NVIDIA has launched Ising, a set of open AI models designed for quantum computing research.

Ising focuses on two tasks: calibration and decoding. Calibration helps tune quantum hardware so that it behaves correctly, while decoding is linked to error correction, which is essential because quantum systems are highly unstable and prone to noise.

Quantum leap:

• Unlike large language models such as ChatGPT, Ising is not meant for conversation or general reasoning. It is a specialised model trained to understand and improve quantum systems. The Ising models target quantum error correction and calibration. 
• “AI is essential to making quantum computing practical,” said Jensen Huang, Founder and CEO of NVIDIA. “With Ising, AI becomes the control plane—the operating system of quantum machines—transforming fragile qubits to scalable and reliable quantum-GPU systems.”
• The launch is part of NVIDIA’s broader push into domain-specific AI models. Ising joins a growing catalogue of open models within NVIDIA’s ecosystem, such as Nemotron for agent-based AI, Isaac GR00T for robotics and BioNeMo for life sciences.

News & updates

• Chips deal: Meta and Broadcom have announced an expanded multi-year partnership to co-develop multiple generations of custom silicon. The collaboration is a major step for both firms as they seek to build the massive computing foundation required for modern AI at a global scale.
• Brand ambassador: Infosys has announced a multi-year partnership with tennis player Carlos Alcaraz. The 7-time Grand Slam champion will be Infosys’ global brand ambassador. Infosys will work with Alcaraz and his coaching team to develop an AI-powered match analytics and a personalised performance application. There will also be joint initiatives between Infosys and the Carlos Alcaraz Foundation. 
• Economic growth: US Treasury ​Secretary Scott Bessent said that the US economy will be slower this quarter amid the ​Iran war, but it ​is in good shape and ⁠will rebound. He also said that ​oil prices do not appear ​to be weighing on inflation expectations.

Which company was originally incorporated as Cadabra Inc.?

Answer: Amazon

We would love to hear from you! To let us know what you liked and disliked about our newsletter, please mail [email protected]. 

If you don’t already get this newsletter in your inbox, sign up here. For past editions of the YourStory Buzz, you can check our Daily Capsule page here.