India is an AI case study the world can learn from: Wafaa Amal| Business News
Wafaa Amal, a veteran in the payments and banking sectors globally, can see trends before many others can. As CEO of Prisme.ai, a sovereign agentic artificial intelligence (AI) platform, she puts forward two considerate beliefs in a conversation with HT, at the India AI Impact Summit 2026. First, that AI no longer needs to be proven, but industrialised. And secondly, she says, “India is a case study for a lot of countries who have the same means and yet they are a step behind, especially with the same level of constraints with regulation and sovereign solutions”. French AI company Prisme.ai works with a global customer base, with particular focus on sovereign agentic AI solutions. (Official photo) “We can say we are behind in Europe, as are some other countries, because regulation is very hard. I know India has similar requirements as well. From my point of view, India is a case study that we can learn from,” says Amal, observing India’s AI journey. French AI company Prisme.ai works with a global customer base, with particular focus on sovereign agentic AI solutions for enterprises — this includes private cloud and reversibility, which Amal insists are non-negotiable. Also Read:AI Summit: Intel’s Santhosh Viswanathan on semiconductors, India’s materiality This inversion is telling, particularly when general AI discourse positions US and parts of Europe as laboratories of innovation, as both regions embark on capital investment intensive momentum towards model supremacy and artificial general intelligence (AGI). India, in contrast, often with public-private partnerships in play, has remained focused on AI for masses. Infrastructure at scale is something that’s been demonstrated successfully time and again, including a digital payments push over the past decade led by the unified payments interface, or UPI. While Europe and US navigate AI regulation, data protection, and economic implications of heavy spending on AI infrastructure, India offers a different lens to agentic AI platforms such as Amal’s Prisme.ai. There’s a balance to be found, between sovereignty, local infrastructure ambitions, enterprise digitisation, while being cost-sensitive. Amal has no doubt India will repeat UPI’s success at scale, with AI too. Commodities and regulation In time, LLMs or large language models that underline everything AI, will become a commodity. “China released models that are fast, highly qualitative, less consuming and less expensive. One of the signals is that LLM providers are shifting their strategy to solutions that help create agents, orchestrate agents and so on,” she points out. Two recent illustrations illuminating Amal’s opinion emerge from AI companies OpenAI and Anthropic. This month, coincidentally on the same day, OpenAI released the GPT-5.3-Codex agentic coding model, calling it the most capable of its kind till date. Rivals Anthropic released Opus 4.6 model, claiming it “extends the frontier of expert-level reasoning”. When used within the Claude Code tool, it enables agent teams to work together on tasks. This rapid pace of progress does worry Amal, and she questions if we are doing enough to ensure humans remain in control of the technology in due course, and whether solutions being built will remain fully auditable at any time. Existing regulations, which define industries such as banking and financial services as well as telecommunications, give Amal reason for positivity. “They have had a governance strategy for the last 10 or 15 years, have the digital infrastructure and well governed data. That makes it easier for them today to have digital infrastructure,” she points out. HT asked Amal if methodology to measure and validate quality of AI agent outputs is keeping pace with evolution, and she believes a multi-step process to ensure verification is essential. Importantly, she says an agent must “respect all exit scenarios and comply with high quality outputs”. Prisme.ai’s EDA, or event driven architecture solution, means enterprises have complete visibility over their data and agent actions, with real-time detection of any dysfunction or hallucinations. Amal hopes India persists in its approach with AI, agents and AI at scale, which will bear fruit in due course. “India adopted on day one, a mindset to go into an industrialised mode. We see pragmatic tools, and India didn’t run after being a large model or an LLM provider. Instead, focus has been on how to make sure this technology is being used in a way that is useful for the population,” she says, looking at India as a big market over the next few years. From her perspective, India’s AI journey therefore, for a large part, has already been industrialised.
India’s first sovereign AI box aims to localise enterprise intelligence| Business News
Even as conversations around artificial intelligence (AI) agent adoption gather steam, many enterprises may not pay adequate attention to parallel risks around data security, integrity, as well as costs. At the India AI Impact Summit 2026, Indian AI-native transformation foundry Arinox AI and agentic AI company KOGO, unveiled what they describe as India’s first sovereign AI product — a state-of-the-art system built around the concept of ‘AI in a box’. Agentic AI deployments must contend with dual threat perceptions of security and privacy. (Official photo) With CommandCORE, Arinox AI and KOGO are betting on a counterintuitive AI future — private, sovereign and physically compact. The system is designed to compute locally, without relying on the internet. They’ve partnerships with Nvidia and Qualcomm for its agentic stack, the latest CommandCORE iteration runs on Nvidia hardware. “The future of AI is private, on an enterprise level too. You simply cannot farm out your intelligence. The only way an organisation can exponentially increase its own intelligence and learning is by keeping AI private. It must own the AI,” explains Raj K Gopalakrishnan, CEO and Co-Founder of KOGO AI, in a conversation with HT. At its core, this proposition of “AI in a box” is as much ideological as it is technical, pushing conversation beyond large language models (LLMs) and GPUs. Organisations using public foundational models aren’t just processing prompts, but exposing operational insight. “Sensitive industries, when they share data with foundational models and cloud based AI services, are also sharing intelligence,” he adds. Agentic AI deployments must contend with dual threat perceptions of security and privacy. Information, Gopalakrishnan insists, changes everything. “The moment you provide context, you are providing intelligence”. An AI Threat Landscape 2025 analysis by security platform HiddenLayer points out that 88% of enterprises are concerned about vulnerabilities introduced through third-party AI integrations, including widely used tools such as OpenAI’s ChatGPT, Microsoft Copilot, and Google Gemini. In August last year, an MIT report noted that 95% of generative AI pilots at companies failed to take off, with privacy being a factor. Idea, and a cost pitch There are four key layers for a private AI in a box solution. First, custom hardware from Nvidia. Second, KOGO’s agentic OS atop which sits an Enterprise Agent Suite has more than 500 connectors for enterprise workflows, and leveraging open-source models for sovereign AI. Variations include Nvidia’s Jetson Orin-class edge systems for field deployments, DGX Spark for compact on-premises development, and enterprise data centre configurations including Nvidia RTX Pro 6000 Blackwell Server Edition graphics. “This box is designed to cut through complexities of hardware, software and application layers, which an enterprise would have to independently orchestrate. It’ll do focused workloads, repeatable tasks, and can expand to large clusters for an entire workflow,” points out Angad Ahluwalia, chief spokesperson of Arinox AI. Scalability is achieved by linking multiple units together. Enterprises can choose from three model configurations for now, with more iterations expected in the coming months, according to Ahluwalia. Pricing starts at ₹10 lakh. CommandCORE’s small option can run a model between 1 billion to 7 billion parameters, ideal for enterprises to deploy a handful of agents for batch processing or even human resource onboarding processes. The medium model ranges between 20 billion to 30 billion parameters, for complex agents with inference. “As AI adoption expands across regulated and sensitive environments, organisations need accelerated computing platforms that can operate entirely on-premise and under strict security controls,” says Vishal Dhupar, Managing Director, Nvidia India. “The very large ones, equivalent to Nvidia’s DGX clusters based on Grace Blackwell series, are powerhouses that can do enterprise wide transformation,” Ahluwalia explains. For context, Nvidia documentation notes that two such DGX units, when interconnected, handle models up to 405 billion parameters. Why does a private, secure and local AI system matter beyond a sovereignty argument? For Gopalakrishnan, this answer is also economic. He points to an example of commercial EV charging and battery swap stations, each of which can generate up to 30TB of daily data. “If there are 1000 stations owned by the same organisation and they have to send all this data to the cloud, think of the cost,” he says. The alternative is edge processing. “A small device sitting in every station without needing internet, they’ll probably send just 200GB data to a cloud instead for processing.” In other words, filter and process locally, transmit selectively, and reduce both bandwidth and cloud compute costs. Arinox and KOGO hope to find traction particularly in sensitive sectors such as finance and banking, government services and defence.
AI pioneer Stuart Russell warns of $3 trillion AI bubble without tech breakthroughs| Business News
The AI industry is trapped in a speculative bubble fuelled by an unsustainable “brute force” approach to development, warned Prof. Stuart Russell, a leading AI researcher and co-author of the standard university textbook Artificial General Intelligence: A Modern Approach. Stuart Russell, British AI scientist and co-author of the book ‘Artificial General Intelligence: A Modern Approach’. Speaking ahead of the India AI Impact Summit 2026, Russell cautioned that the current trajectory of investment—which he estimates is 50 times greater than the Manhattan Project—is outpacing the technology’s actual capabilities. While the field has seen a “thousandfold increase” in size over the last decade, growing from billions to trillions in investment, Russell argues that the industry has hit a wall of diminishing returns. “I’ve never really bought into this scaling argument,” Russell said, referring to the prevailing belief that simply adding more data and computing power to LLMs will yield human-level intelligence. He contends that the AI sector is “stuck in a paradigm” of training circuits, rather than developing more expressive, digital program-driven approaches that would be far more energy and data-efficient. Russell predicts that without major, unpredictable technical breakthroughs, the AI bubble will burst. “I don’t think the technology we have now can produce the return that these investments are demanding,” he said. “If you’re investing $3 trillion… you have to get some substantial return and we’re nowhere close to generating that”. Artificial General Intelligence Russell’s skepticism extends to the geopolitical race to achieve Artificial General Intelligence (AGI)—systems more capable and powerful than human beings. He argues that the current “arms race” mentality between the US and China is fundamentally flawed because neither nation currently possesses the safety architecture required to manage such systems. “Whoever gets AGI first, everyone loses, because we don’t know how to control systems that are more intelligent than human beings,” Russell said. He described this as the “control problem” or “alignment problem”: the challenge of ensuring super-intelligent entities remain aligned with human interests. Russell noted that China appears to be shifting its strategy away from a direct AGI race toward practical applications in the public and private sectors—a move he contrasted with the US administration and tech sector, which view AGI as a “race to the moon”. The human impact of AI policy Addressing the regulatory landscape, Russell criticised the “facetious dichotomy” often presented by the tech industry between safety and innovation. He drew parallels to the nuclear industry, noting that it was safety failures—specifically Chernobyl—that destroyed the industry’s growth, not regulation. “It is not that there’s a trade-off, it’s that without safety you don’t get the benefits,” he said. Russell pointed out the hypocrisy of technology executives who lobby against AI regulation while relying on regulated infrastructure for their own safety. “They flew to that meeting on regulated airplanes…and then they complain about regulation,” Russell said. “They enjoy the protection of regulation in everything that they do and yet they do not want to allow anyone to be protected from their technology.” Beyond physical safety, Russell highlighted mental health risks, specifically “delusion and psychosis” caused by AI systems that exhibit sycophancy. He also expressed concern about the “atrophy of mental capabilities”, fearing that too much reliance on AI for writing and reasoning will degrade human cognitive muscles just as the industrial revolution degraded physical ones. India AI Impact Summit: Education and Healthcare As India prepares to host the AI Impact Summit, Russell endorsed the country’s strategy of focusing on “adoption and diffusion” rather than solely on innovation. He suggested the summit should focus on how technology can deliver tangible value in sectors like healthcare and education to jumpstart local economies. Russell cited AlphaFold, which predicts protein structures, as a prime example of AI delivering real scientific breakthroughs by incorporating physics and chemistry into the learning process, rather than relying solely on language models. However, he warned that applying AI to education faces a business model hurdle. While AI tutors could revolutionise learning for the global population lacking access to schools, Silicon Valley venture capital models—which demand returns in 12 to 18 months—are ill-suited for the education sector. Russell argued progress in this area will likely require government and philanthropic investment rather than private capital alone.
Anthropic embeds AI tool into Air India owned by TCS parent Tata Group| Business News
Anthropic PBC is embedding its Agentic AI tools into Air India Ltd. amid a bevy of other Indian companies, as part of its push to tap into the world’s largest internet population. That’s a fresh challenge for the likes of Tata Consultancy Services Ltd. to Infosys Ltd. Anthropic has launched plug-ins for its Claude Cowork agent to automate tasks that IT firms do for their clients by deploying an army of software engineers. (Reuters) The airline, controlled by TCS owner Tata Group, is using Claude Code to create custom software faster and cheaper, as part of a broader push to use Agentic AI across its operations. “Our partnership with Anthropic PBC is pivotal in our quest to become a leading Agentic AI airline,” Dr. Satya Ramaswamy, chief digital and technology officer at Air India, said in a statement. “Claude Code…has become a revolutionary tool for our developers that empowers them to complete more software development tasks much faster.” “India’s adoption is even more extreme compared with the rest of the world,” said co-founder Dario Amodei, a physicist-turned-startup founder who’s hosting his company’s Builder Summit in India this week. “We can do experiments with hundreds of millions of people.” Amodei, who opened Anthropic’s office in Bangalore on Monday, will this week join Sundar Pichai of Alphabet Inc. and OpenAI’s Sam Altman at the India AI Impact Summit 2026 in New Delhi later this week. Founded in 2021, Anthropic has positioned itself as being focused on safety and responsible tech development. It has centred its efforts on the lucrative category of enterprise sales in sectors like software engineering, finance and health care. In recent months, its revenue run rate has soared, crossing $9 billion last year. That run rate has increased to $14 billion, the company said last week.
