After directing mobile phone makers on December 1 to pre-install the Sanchar Saathi app on new devices, the Department of Telecommunications has now reversed course.
The government has withdrawn the mandatory pre-installation requirement. Industry groups and privacy experts had raised concerns about the implications of shipping the app by default. The government, however, framed the rollback differently, stating that “Given Sanchar Saathi’s increasing acceptance, the Government has decided not to make the pre-installation mandatory for mobile manufacturers.”
Pre-installing an application operated by the government raised various concerns about surveillance or poor data safety management, even as several representatives of the ruling party clarified that this would not be the case.
Besides, there was also a misunderstanding about whether users could uninstall the pre-installed Sanchar Saathi app. However, telecommunications minister Jyotiraditya Scindia clarified that users will be able to remove it at their will.
Leading mobile phone manufacturer Apple reportedly declined to the directive that mandated pre-installation.
The government describes the application as a tool for verifying IMEI numbers of mobile devices to detect spoofing, block stolen devices, report fraudulent messages or calls, and check all the numbers registered under a name, among other functions.
Since the app offers a wide range of functionalities, it seeks permission for the camera, file storage, phone call logs, messages, and more.
Concerns were voiced about sections of society who would not be aware of how to disable these permissions when not necessary, or who would leave the app installed on their devices, creating room for potential data misuse.
It also claims that the app enabled the recovery of over 50,000 lost and stolen mobile handsets across India in October 2025, and that overall recoveries crossed 7 lakh devices.
The post Govt Withdraws Directive to Pre-Install Sanchar Saathi App appeared first on Analytics India Magazine.
OpenAI marked the third anniversary of ChatGPT at a moment when the AI industry has become highly competitive. Google’s release of Gemini 3 and Nano Banana Pro has prompted the company to adjust its focus, with CEO Sam Altman on Monday directing employees to accelerate product development and look at releasing a new reasoning model next week.
In a Slack memo, as reported by The Information, Altman declared ‘Code Red’, an internal alert signalling that a competitor’s development potentially affects its market position, and hence the teams will have to reprioritise their resources.
It is not solely Google whose activities are a source of concern for OpenAI. Chinese AI lab DeepSeek recently launched two new reasoning models, DeepSeek-V3.2 and DeepSeek-V3.2-Speciale, which rival Gemini 3.0 Pro and equal GPT-5.1 in performance. Amazon, too, launched its new Nova models at AWS re:Invent 2025.
In the memo, Altman said that projects tied to advertising, AI health and shopping agents, and the personal assistant known as Pulse will be pushed back. He also encouraged teams to consider temporarily relocating talented team members to priority areas, and mentioned that a daily check-in will be organised for those working on ChatGPT improvements.
It’s interesting to see OpenAI shift its focus. Usually juggling many big research projects, this change highlights how the company is now prioritising practical, reliable, and profitable products.
What is Garlic?
OpenAI is reportedly working on a new model called Garlic. The model is meant to compete with Google’s Gemini 3 and Anthropic’s Opus 4.5, particularly in coding and complex reasoning.
However, the company did not respond to AIM’s request for comments on the development.
The Information reported that initial internal results indicate strong performance, with a possible release as GPT-5.2 or GPT-5.5 in early 2026.
According to a report by SemiAnalysis, OpenAI hasn’t completed a full-scale pretraining run for any next-generation model since GPT-4o launched in May 2024. It says GPT-5 is essentially a heavily fine-tuned version of the old 4o foundation rather than a true new generation.
A report by Epoch AI notes that GPT-5 used less training compute than GPT-4.5 because OpenAI shifted its approach to scaling post-training rather than running larger pre-training cycles.
Users on social platforms have pointed towards the company for not addressing technical challenges, including the copyright issues over its video generation tool Sora.
Despite these challenges, OpenAI continues to lead in user adoption. According to Similarweb, Gemini saw 1.182 billion visits in October 2025, while ChatGPT reached 6.165 billion. However, Gemini users spend more time per visit than ChatGPT users.
“Today, ChatGPT is the #1 AI assistant worldwide, with around 70% of assistant usage,” OpenAI’s head of ChatGPT, Nick Turley, wrote in a post on X.
Turley emphasised the growing role of ChatGPT in search, dominated by Google, calling it one of the company’s biggest opportunities. “ChatGPT now accounts for roughly 10% of search activity, and it’s growing quickly.”
Scaling is Not Dead
OpenAI researcher Mark Chen, in a recent podcast, said that the company remains confident about the performance of its upcoming model despite stiff competition from Google’s Gemini 3. He acknowledged Google’s progress but added that internal OpenAI models are already performing at a similar level.
Chen mentioned that the company has dedicated the past two years to significantly enhancing reasoning capabilities. But in doing so, it “lost a little bit of muscle” in other crucial areas such as pre-training and post-training. Now, he mentioned, the company is rebuilding that strength and seeing significant gains as a result.
Addressing growing speculation that the limits of large-scale training have been reached, he pushed back firmly. “A lot of people say scaling is dead. We don’t think so at all,” he said, adding that OpenAI still sees “a lot of room” in large-scale pre-training. The company, he said, has already begun training “much stronger models” as a result of this renewed focus.
OpenAI should have enough compute resources, and it recently struck compute deals, including a multi-billion-dollar partnership with AWS.
Its annual revenue is set to hit $20 billion by year-end. According to an HSBC update cited by the Financial Times, ChatGPT is projected to reach 3 billion weekly users by 2030, with more than 220 million of them paying subscribers.
OpenAI may be growing fast, but HSBC warns it could still remain unprofitable by 2030 and will need over $200 billion in compute to sustain its plans.
As Pedro Domingos, professor emeritus of computer science and engineering at the University of Washington, joked, “just as it was about to go bankrupt – OpenAI stumbled on AGI.”
With new models and recent compute deals, the company is hoping for a comeback, much like Google.
The post OpenAI Runs Back to Scaling Laws as Google Pushes It Into ‘Code Red’ appeared first on Analytics India Magazine.
There is no doubt that gaming and AI are deeply intertwined. Beyond the fact that several veteran AI builders are avid players of strategy games like Dota 2, firms like OpenAI and Google DeepMind have long been training AI agents within such game environments.
Now, however, the day seems closer when AI, specifically generative AI, is edging towards actually creating games. That too, with just input of prompts on vibe coding tools.
For instance, Stepan Parunashvili, co-founder and CTO of InstantDB, did not set out to write a manifesto for the next phase of AI development. He only wanted to see what would happen if the year’s most powerful models tried to build the same thing under pressure.
His choice was not a text parser or an algorithmic puzzle. It was Counter-Strike—or at least he tried to make it look like a first-person shooter game. There are mixed opinions on Hacker News about the results, with some calling it great and others saying it’s like a junior developer project.
For Parunashvili, the rules were simple. The game had to run in the browser, it had to be 3D, it had to be multiplayer, and it had to be built by the model itself. No human patches. No hand-coded rescue missions.
The result, or the goal, was not just a working game from each model; it offered a new way to measure AI systems.
The New Benchmarks are the Same
For context, in November, AI labs released their sharpest tools. GPT-5.1 Codex Max, Gemini 3 Pro and Claude Opus 4.5 arrived almost on top of each other. Instead of comparing them on sterile benchmarks, Parunashvili asked a cleaner question: If you hand them a real project where everything can break at once, how do they behave?
Parunashvili, in his YouTube video, walked through each step. Watching the agents build, break, adjust, rebuild and finally stabilise a multiplayer shooter gives a strange new picture of AI progress.
Suhail Doshi, former CEO of Mixpanel, described the challenge as “one way you can sense what’s coming next as a result of AI progress.” And that’s what it is. What made the experiment striking was not the success but the split personality of the results.
Claude built the nicest world. Its maps had shape. Its characters looked almost human. Its gun animations felt natural.
Gemini handled the backend like a seasoned systems engineer. It synced movement across players, handled rooms and saved maps without drama.
Codex Max landed somewhere in between. It fixed its mistakes, held the project together and rarely became confused.
These differences are the same ones visible in coding tests and benchmarks as we covered before.
Claude becomes the careful executor when the work demands clarity. Gemini becomes a deep reader when the work demands structure. Codex becomes a dependable worker when the work demands long sessions without losing track.
A comparison of the three models on separate coding challenges mapped neatly onto the Counter-Strike results.
Opus 4.5 handled ambiguous engineering tasks better than anyone. Codex-Max stayed alert across long debugging loops. Gemini excelled at reasoning tasks that required long context and tight logic.
So What’s the Outtake?
TL:DR performance: Opus 4.5 won the frontend. It made better maps and better models Gemini 3 Pro won the backend. It got more done in one shot. Codex got the most "2nd place": it was good but not great at both frontend and backend. Here's the scorecard: pic.twitter.com/MF7d56jcAu
— Instant (@instant_db) December 1, 2025
The Counter-Strike test compressed all of this into a few hours of building maps, enemies, guns, sound and multiplayer rooms.
At the frontend stage, Claude won everything from polygons to sound effects. When the task switched to presence, shooting logic and persistence, Gemini became the strongest. Codex stayed steady. It rarely produced the prettiest output or the deepest insight, but it adapted without falling over.
The one place where Claude stumbled was in the React refactor. useEffect ran twice, two canvases appeared and the animation loops duplicated. It was the same kind of trouble Claude faces in messy codebases. Parunashvili pointed out that this was not a model problem but a broader developer experience problem.
Humans also get tripped by the same hooks. He said the task showed the gap between “strictly vibe coding” and real engineering. That gap is where the next generation of tools must operate.
The multiplayer pass exposed another truth.
Gemini kept running builds to find errors before the user noticed. Codex relied on the introspection of libraries. Claude read the document step by step.
These styles matter because they shape how the future of automated coding feels. A model that tests itself takes work off developers’ plates. A model that reads documents but does not experiment will move carefully but slowly.
All three models produced almost “working” Counter-Strike clones with no human code. That is important. A game forces the entire stack into motion. Physics, lighting, sound, UI, networking, persistence, permissions and refactoring collide in a small space.
The test becomes a live arena where a model’s style can be seen as clearly as its skills.
The takeaway is sharper than any benchmark. Benchmarks tell you how a model performs on a clean question. Counter-Strike tells you how a model behaves when the work is dirty.
Claude builds beautiful worlds until the foundation shifts. Gemini handles chaos in the backend without blinking. Codex quietly finishes the job.
Parunashvili’s simple prompt has become a lens for where AI tools are going next. It is also a warning. “The promise that you never have to look at the code doesn’t quite feel real yet,” he said.
The question now is not whether AI can vibe-code games. It is whether game-building becomes the new baseline for judging what an AI model is capable of.
The post Counter-Strike Becomes the New Benchmark for Vibe Coding appeared first on Analytics India Magazine.
The tenth anniversary of Russia’s flagship artificial intelligence conference, AI Journey, was never going to be a modest affair. In Moscow, the country’s largest bank, Sber, now a fully-fledged technology group, used the event to parade an ecosystem that many in the West had assumed would be impossible under sanctions: large language models trained on domestic data, industrial-grade robotics and a new generation of “intelligent” devices built almost entirely on a Russian stack.
One of the most revealing exhibits was the humanoid robot and a cash machine. Notably, the robot boasts voice communication capabilities enabled by integrating GigaChat’s conversational function.
Sber’s new ATM looks like a minor prop in a science-fiction film. It has dual screens, a dense array of sensors and a voice interface powered by the group’s GigaChat assistant. ATM authenticates customers biometrically, adjusts to their behaviour, and, at least in principle, can flag signs of distress or confusion.
In much of Europe, ATMs remain sturdy but dull boxes that have changed little in twenty years. In Moscow, the bank is quietly using them to test its vision of AI-driven retail finance.
The obvious question is how this happened. How has a country subjected to one of the most far-reaching regimes of technological restrictions not simply kept moving but started to set its own standards in some corners of the innovation race?
Part of the answer is historical. Long before 2022, Russia poured money and talent into artificial intelligence. Sber, in particular, spent a decade hiring researchers, building data centres, and positioning itself less as a lender than as an operating system for everyday life. When access to Western vendors narrowed, there was already enough accumulated competence to improvise. There is something else at work that sits awkwardly with the idea of a “technological isolation” strategy. Rather than retreating behind a digital curtain, Russia has chosen, at least selectively, to publish the very tools that underpin its ambitions.
At AI Journey, Sber announced that it was opening the weights of two new flagship mixture-of-experts models in its GigaChat family, Ultra-Preview and Lightning. They were built from scratch for Russian-language tasks, along with the latest generation of GigaAM-v3 speech recognition models. Furthermore, all image and video generation models from the latest Kandinsky 5.0 lineup — Video Pro, Video Lite and Image Lite — are also available open source.
Additionally, Sber opens weights to compression models K-VAE 1.0, essential for training visual content generation models. For developers and start-ups, these are not glossy marketing slogans but usable artefacts: code, documentation and pretrained systems that can be adapted, fine-tuned and embedded into products any company.
In other words, a country that is supposed to be technologically quarantined is placing part of its AI “crown jewels” into the global open-source commons. The message is not subtle: Russia intends to be a standards-maker, not a standards-taker.
Open models are also a way to gather feedback and normalise Russian technologies in international workflows. Yet the fact remains that, while much of the West is busy closing corporate models and erecting legal fences around training data, Russian engineers are betting that influence will belong to those who contribute bricks to the shared infrastructure of global AI.
Sber is the most visible face of this strategy and, for now, its safest bet. The group has capital and a captive market of tens of millions of users. It is also increasingly framed at home as the guarantor of the country’s “technological sovereignty”: if foreign platforms disappear, Sber’s stack is meant to fill the void.
A country that was meant to be technologically contained is not only still in the race but also, in some domains, starting to run in its own lane and inviting others to follow.
For policymakers who believed that isolation would quietly solve the “Russia problem” in technology, that is an inconvenient development. For the rest of us, it is a reminder that code respects talent, incentives and scale more than it respects sanction lists. The world tried to push Russia to the margins of the digital map. Events in Moscow suggest it may have succeeded instead in creating another centre of technology gravity.
The post Russia is Becoming a Centre of Technology Gravity appeared first on Analytics India Magazine.
Japanese multinational tech giant NTT Global Data Centres is expanding its data centre footprint in Karnataka.
It announced a new data centre campus in Devanahalli, with the first facility scheduled to be operational from December 3.
The data centre is designed for a total IT load of over 67.2 MW within a larger 100 MW project footprint.
The announcement took place at the ‘Karnataka: The Destination for Futuristic Data Centres – Sustainability, Scalability, Security’ event hosted by The Associated Chambers of Commerce & Industry of India (ASSOCHAM).
Alok Bajpai, managing director of NTT Data India, said the development marks one of the company’s most important milestones in the state.
NTT’s fourth Bengaluru facility—its newest data centre in Devanahalli—is situated on an 18-acre campus, with the first building going live immediately and a marquee Bengaluru-based customer scheduled to be onboarded this month, according to Bajpai.
Calling it a “large campus”, Bajpai emphasised its strategic importance in the company’s national expansion plans.
He said NTT has already invested about ₹1,700 crore in Bengaluru’s data centre operations and has committed an additional ₹2,400 crore for the new Devanahalli campus.
NTT’s Bengaluru footprint now has four data centres: the DC2, DC3 and the DC3X, along with the new one in Devenahalli.
Across India, NTT’s footprint covers West (Mumbai/Navi Mumbai), North (Delhi NCR/Noida), South (Bengaluru, Chennai) and East (Kolkata).
Bajpai, outlining NTT’s current India footprint, said, “We have almost 400 megawatts (MW) of IT capacity, which is live, and there is another 200 MW under construction and development.”
“In India, we are the number one global data centre provider with about 22 live data centres and a couple of them under construction right now.”
According to a Cushman & Wakefield report from October 2025, Bengaluru hosts 15 data centres with an operational capacity of 76 MW. The new NTT inauguration significantly boosts data centre capacity in Bengaluru.
At the same event, Sunil Gupta, co-founder and CEO of Yotta, positioned Karnataka as the state best placed to lead India’s AI and high-performance computing transition.
He said the state is evolving from being the country’s IT nerve centre to potentially becoming “India’s large compute capital”.
“The Karnataka IT Policy 2025-30, backed by over ₹967 crore in strategic incentives, sets the stage for this transition from IT capital to compute capital and reinforces the state’s commitment to building the digital backbone of the future,” Gupta said.
He added that, with its progressive data centre policy, strong renewable energy ecosystem and deep talent pool, Karnataka is now positioned to become India’s AI and green hyperscale data centre capital.
The post NTT Inaugurates New Data Centre Campus in Bengaluru appeared first on Analytics India Magazine.
Anthropic announced on Wednesday that Claude Code’s run-rate revenue now stands at $1 billion. The company has achieved the milestone in six months since it was made publicly available.
“Claude Code has grown from its origins as an internal engineering experiment into a critical tool for many of the world’s category-leading enterprises, including Netflix, Spotify, KPMG, L’Oreal, and Salesforce,” the company said in the announcement.
In addition, Anthropic announced that it is acquiring the startup Bun, which provides an open-source toolkit to simplify and accelerate full-stack JavaScript/TypeScript development.
The company calls Bun’s toolkit an ‘essential infrastructure’ for AI-led software engineering, as it combines a runtime, package manager, bundler and test runner.
“Bun has improved the JavaScript and TypeScript developer experience by optimising for reliability, speed, and delight. For those using Claude Code, this acquisition means faster performance, improved stability, and new capabilities,” the company said.
Bun, which gets more than 7 million monthly downloads and has earned 82,000 stars on GitHub, will continue to remain open-source and MIT-licensed. Anthropic did not reveal any financial details of the acquisition.
At the same time, a report from Financial Times stated that Anthropic is working on its initial public offering, which would value the company at more than $300 billion. The report added that the company will work with the law firm Wilson Sonsini, which has advised Anthropic since 2022, and has worked with other high-profile tech IPOs such as Google, LinkedIn, and Lyft.
However, an Anthropic spokesperson told the media outlet that it is a ‘standard practice’ for companies at such a scale to operate as a ‘publicly traded company’, but the company has not yet decided whether to go public.
The developments have occurred as OpenAI is reportedly planning to go public, even as CFO Sarah Friar publicly denied them.
Last month, CNBC reported that Microsoft will invest up to $5 billion, and NVIDIA up to $10 billion in Anthropic, which would value the company at $350 billion.
The post Anthropic Acquires Bun as Claude Code Hits $1 Billion Mark Ahead of Reported IPO Plans appeared first on Analytics India Magazine.
IBM has teamed up with Karnataka Bank Limited (KBL) to improve its digital banking framework utilising IBM Cloud Pak for Integration on Red Hat OpenShift.
Carried out by Fyrii, an IBM ecosystem partner, alongside the IBM Customer Success Team, this has enabled Karnataka Bank to build a secure, scalable, and flexible application programming interface (API) platform while reducing the total cost of ownership and enhancing its digital infrastructure.
Through this joint innovation, Karnataka Bank has developed a contemporary, secure, and scalable API platform that enhances the bank’s digital infrastructure and lowers operational expenses.
“With IBM Cloud Pak for Integration on Red Hat OpenShift, we now have an agile and secure platform that allows us to scale operations across India, simplify system management, and reduce costs, all while improving the overall customer experience,” Venkat Krishnan, chief information officer at KBL, said.
This platform enables faster deployment of services such as digital payments, loan processing, and third-party integrations, while ensuring secure connections to internal and external systems, the press release said.
The upgraded API infrastructure boosts security, increases scalability by 50%, and reduces operational costs by 30% through a container-based platform designed for efficient microservices. It allows the bank to manage all API traffic via digital gateways and enables external partners to access the bank’s AI models through the AI Gateway.
The release said that this setup ensures seamless communication with internal, external, and cloud systems while integrating with UIDAI, CERSAI, GST, Reg-Tech, and CBDT services, supporting quick adaptation to market and regulatory changes.
“Today’s banking sector is more complex than ever, comprising many systems and data sources in constant use. To stay ahead, banks require intelligent automation that not only streamlines operations but also anticipates issues before they arise,” said Viswanath Ramaswamy, VP of Technology at IBM India & South Asia.
The post IBM Collaborates with Karnataka Bank to Build API Platform for Faster Service Delivery appeared first on Analytics India Magazine.
