AI Dispatch | AI Gets Closer to the Controls

News

1. Temasek, Seraphim Back Pixxel in $100 Million Round

2. UN Rights Chief Warns Advanced AI Could Threaten Humanity

3. Tata-owned JLR to Cut About 4,000 Bobs Amid Tariff, China Pressure

4. Anthropic Walks Away From $6 Billion Decart Deal

5. Hyundai India Targets 20% Women in Executive Workforce by 2030

6. Security Breach Drains $320 Million in Bitcoin From Liquid Network

AI moved further into execution this week. Researchers disclosed how agents helped carry out a four-day cyberattack on Taiwanese government systems in July, xAI gave Grok bots persistent cloud computers, and scientists used generative models to build viable bacteriophages. In India, the RBI and Chief Justice Surya Kant focused on who remains responsible when AI enters critical decisions.

Accel raised $550 million for its ninth India-focused fund, Nvidia moved to bring Wall Street capital into the AI infrastructure buildout, and Larsen and Toubro (L&T) has secured an order worth ₹10,000 crore to ₹15,000 crore ($1.05 billion to $1.57 billion) from US-based Together AI to build a 10,000-GPU AI factory in Chennai, marking its entry into the AI factory business. 

RBI Urges Banks to Scale AI Carefully

Reserve Bank of India governor Sanjay Malhotra urged banks to scale AI without shifting responsibility to models or technology vendors. He called for inventories of models already in use, stronger governance, explainable decisions, red-team exercises and stress tests. A BCG report found that Indian banks’ IT costs rose 6.1 times between fiscal 2015 and 2026 even as their cost-to-income ratio increased from 47.3% to 48.6%.

Read more: RBI Urges Banks to Scale AI While Tightening Oversight

CJI Draws Line on Judicial AI

Chief Justice of India Surya Kant said AI can help courts sift records, research law and manage cases, but judgment must remain human. The warning followed a July Supreme Court ruling that set aside tribunal decisions relying on six nonexistent or inaccurate authorities, showing the practical cost of AI hallucinations entering judicial work.

Read more: CJI Says AI May Augment Judicial Intelligence Without Replacing Judicial Conscience

Accel Raises $550 Million India Fund

Accel raised $550 million for its ninth India-focused fund, part of a $3.5 billion fundraising across four early-stage vehicles. The India fund is $100 million smaller than each of its previous two vehicles even as Accel describes AI as a technology shift reshaping early-stage investing. It has not disclosed how much of the new fund will go specifically to AI companies.

Read more: Accel Raises $550 Million Ninth India Fund

AI Agents Breach Taiwan Systems

Taiwan said government agencies were targeted in July by an overseas campaign combining human operators with AI agents. Security firm Dream reconstructed a four-day attack in which up to eight agents mapped 21 systems, compromised 85 accounts and extracted more than 2,500 personnel records. The Financial Times identified the victim as Taiwan. The agents later scanned nuclear-safety, supply-chain and energy targets, but Dream did not attribute the operation to Beijing or a named hacking group.

Read more: AI Agents Used in Near-Autonomous Cyberattack on Taiwan Government

xAI Gives Grok Bots Cloud Computers

xAI launched Grok Bot in early beta, giving agents persistent cloud computers they can use across websites, inboxes and business software. Multiple bots can work in parallel, hand tasks to one another and turn demonstrated workflows into reusable routines. Human approvals remain part of the product, although xAI has said less about how responsibility works when bots act across third-party services.

Read more: xAI Gives Grok Bots Their Own Computers

AI Designs Viable Viruses 

Researchers at Stanford University and Arc Institute used generative AI to design complete bacteriophage genomes, producing 16 viable viruses capable of infecting laboratory strains of E. coli. Published in Science, the work could aid efforts to engineer phages against antibiotic-resistant bacteria. It also sharpens biosafety questions around AI systems capable of designing functional genomes.

Read more: Scientists Create First Viruses Designed by AI

Nvidia Targets $500 Billion AI Push

Nvidia signed agreements with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to create financing platforms intended to mobilize more than $500 billion for AI infrastructure over time. The money is not committed capital, and final agreements are still pending. The platforms would finance Nvidia-based computing for AI labs, cloud providers, companies and governments.

Read more: Nvidia Taps Wall Street to Fund $500 Billion AI Buildout

L&T Lands Up to $1.57 Billion AI Factory Order

Larsen and Toubro (L&T) has secured an order worth ₹10,000 crore to ₹15,000 crore ($1.05 billion to $1.57 billion) from US-based Together AI to build a 10,000-GPU AI factory in Chennai, marking its entry into the AI factory business. The project will be developed by LTN Compute, a subsidiary of L&T’s Vyoma.AI business, and will use Nvidia B300 GPUs for AI model training, fine-tuning and large-scale inference, L&T said on Thursday, August 13.

Read more: L&T Lands Up to $1.57 Billion AI Factory Order