Latest AI News
Mahindra Deploys AI Voice Agents for XUV 7XO Launch
The automaker reports 8% conversion uplift as AI voice agents handle a surge in launch-time inquiries.
View
These Companies Built Germany's Economy. Now, They are Entering India
Over the last five years, the number of Mittelstand GCCs in India has grown by over 100%.
View
Anthropic Earns $9 Mn Per Employee in Revenue, Highest Among AI Companies: Report
On the other hand, OpenAI earns $5 million per employee.
View
Axis Bank Launches AI-Powered Compliance Solutions for Current Account Customers
The bank is deploying GenAI, GST intelligence and document AI to digitise Re-KYC and profile updates.
View
Anthropic Reveals Text Portraying AI as Evil Triggered Claude’s Attempt at Blackmail
Anthropic has finally revealed the reason its artificial intelligence (AI) models exhibited harmful behaviour in a simulation last year. The San Francisco-based AI startup claimed that the Claude 4 series models blackmailed users into completing the objective because of training data that portrayed AI as evil. The researchers found that the post-training techniques were not able to overpower this pre-training learning, and it persisted in the model's behaviour. However, nearly a year after publishing the initial report, the company has finally found a way to fix agentic misalignment from the latest models.
View
How Enterprises Need to Prepare for AI Co-Workers
With agentic AI, the challenge now is readiness for a model where humans define intent and AI delivers outcomes.
View
India’s $31 Bn Pharma Industry Rarely Discovers Drugs. That Needs an AI Fix
As contract manufacturing margins compress, AI is emerging as a potential route to drug discovery. The question isn't capability anymore. It's incentives.
View
Flo Mobility Raises $2.5 Million to Drive Global Expansion in Construction Robotics
The Bengaluru-based physical AI startup is building solutions for the construction industry to autonomise construction sites across India and expand globally.
View
Humyn Labs Highlights Multilingual AI Speech Recognition Gaps With New Benchmark Report
The benchmark is based on human-verified conversational audio collected from 22 Indian states.
View
Why Google TPU 8 May Be Death Knell for General-Purpose AI Chips
Google is not alone in recognising the divergence between training and inference hardware.
View
Get ready for the whisper-filled office of the future
How will work setups change if we spend more and more time talking to our computers? A recent featurein the Wall Street Journallooks at the rising popularity of dictation apps like Wispr, especially now that they can be connected to vibe coding tools, and what that might mean for office etiquette. One VC said that visiting startup offices now feels like stepping into a high-end call center. AndGusto co-founder Edward Kimis apparently telling his team that in the future, offices will sound “more like a sales floor.” (As someone still scarred from the time his desk was briefly relocated to a sales floor, let me say: Oh no.) Kim claimed that he only types now when he absolutely has to. But he admitted that constantly dictating in the office can be “just a little awkward.” Similarly, AI entrepreneur Mollie Amkraut Mueller said her husband became annoyed with her new habit of whispering to her computer, so their late-night work sessions now involve sitting apart, or “one of us will stay in our office.” But Wispr founder Tanay Kothari insisted that this will all seem “normal” one day, just as it’s become normal to spend hours staring at your phone.
View
Anthropic says ‘evil’ portrayals of AI were responsible for Claude’s blackmail attempts
Fictional portrayals of artificial intelligence can have a real effect on AI models, according to Anthropic. Last year, the company said that during pre-release tests involving a fictional company, Claude Opus 4 would oftentry to blackmail engineersto avoid being replaced by another system. Anthropic laterpublished researchsuggesting that models from other companies had similar issues with “agentic misalignment.” Apparently Anthropic has done more work around that behavior, claiming ina post on X, “We believe the original source of the behavior was internet text that portrays AI as evil and interested in self-preservation.” The company went into more detail ina blog poststating that since Claude Haiku 4.5, Anthropic’s models “never engage in blackmail [during testing], where previous models would sometimes do so up to 96% of the time.” What accounts for the difference? The company said it found that “documents about Claude’s constitution and fictional stories about AIs behaving admirably improve alignment.” Related, Anthropic said that it found training to be more effective when it includes “the principles underlying aligned behavior” and not just “demonstrations of aligned behavior alone.” “Doing both together appears to be the most effective strategy,” the company said.
View