India's undersea turbulence; Anthropic's courtroom win

The government has asked telcos and subsea operators to assess risks to India's data cables amid the war in West Asia. This and more in today's ETtech Top 5.

Also in the letter:

■ IT firms lean on M&A

■ OpenAI's ad pilot mints money

■ Microsoft set for worst quarter

The Centre has asked telecom operators and subsea cable firms to map India's exposure to potential damage to undersea data cables amid the conflict in West Asia and to draw up hard fallback plans, according to people in the know.

Driving the news: The Department of Telecommunications has already met industry leaders as anxiety mounts over Iran's threats to cables that run through the Strait of Hormuz and the Red Sea -- two of the world's most critical data arteries.

Traffic, latency, and cost risks: Roughly a third of India's westbound internet traffic to the US and Europe flows through this corridor. Some of it can be rerouted via Singapore, but capacity there is limited and more expensive. Longer detours across the Pacific would slow speeds further and push up costs, industry experts warned.

Data centre ambitions at risk: Any sustained disruption could derail India's $270-billion data centre push, which hinges on stronger, more resilient subsea connectivity.

Also Read: Google announces $30 million science fund; new subsea connectivity initiative

Dario Amodei, CEO, Anthropic

A federal judge has temporarily barred the US Department of Defense from labelling Anthropic a security risk, giving the AI startup breathing room in its fight over federal contracts and the military's use of its technology.

Breathing room: The ruling is interim, but it allows Anthropic to continue bidding on and executing US government contracts for now.

Also Read: Why the US government labelled Anthropic a 'supply chain risk': A timeline

The case centres on a $200-million contract in which Anthropic sought to bar the use of its models for surveillance and autonomous weapons. The Pentagon has pushed back, stressing it will not let a private company dictate how its technology is deployed.

Company's stance: An Anthropic spokesperson said the firm was "grateful to the court for moving swiftly" and that it "remains focused on working productively with the government to ensure all Americans benefit from safe, reliable AI."

Also Read: Claude Mythos: Leak spills details on Anthropic's new AI model, its most powerful yet

Even as it jousts with regulators, Anthropic is weighing a public listing as early as October, according to reports.

Details:

Also Read: Anthropic Claude can now control your computer: New update explained

Infosys' $560 million investment in Optimum Healthcare IT and Stratus underlines how IT services firms are leaning into specialised solutions and large deals to cushion revenue pressure from AI.

What's going on? India's nearly $300 billion tech services industry could see revenue shrink 2-4% over the next two years as companies accelerate AI adoption. To offset that drag, firms are chasing bigger, more strategic deals and acquisitions.

Large deals:

Also Read: AI software plugins likely to weigh on IT roles for next 12-18 months, experts say

Expert view: Analysts say these deals are coming at attractive valuations and add useful, high-margin capabilities. But despite the spending spree, Karan Uppal, vice president at Phillip Capital, noted that these deals still account for only a modest share of overall revenue.

Muted quarter ahead: Meanwhile, ET reported that fourth-quarter results for India's large-cap IT firms are likely to stay soft, with growth expected between -1.6% and 2%, according to brokerages CLSA and JPMorgan.

Also Read: Nimble mid-tier IT riding large deal momentum as AI rejigs biz Sam Altman, CEO, OpenAI

OpenAI's ChatGPT ads pilot in the United States has already topped $100 million in annualised revenue within six weeks, showing strong early demand for its new ads business.

Quick lookback: In January, Sam Altman-led OpenAI said it would begin testing ads in ChatGPT, with a limited group of users in the US, aiming to fund the heavy costs of building frontier AI models.

Also Read: OpenAI set to raise about $10 billion from MGX, Coatue, Thrive

The ads currently appear to users on the free tier and the lower-priced Go plan.

OpenAI had stressed that ads are separated from ChatGPT's responses and do not affect its outputs, insisting that user conversations are not shared with advertisers.

Also Read: ETtech Explainer: Why is OpenAI shutting down Sora, its text-to-video app?

Tell me more: The company said around 85% of users are currently eligible to see ads. Still, fewer than 20% actually see them on a given day - suggesting headroom to grow ad revenue even before expanding the user base.

OpenAI now plans to roll out the pilot to more countries in the coming weeks, including Australia, New Zealand and Canada.

Also Read: SoftBank secures $40 billion loan to fund further OpenAI investment

Satya Nadella, CEO, Microsoft

Microsoft is under pressure on two fronts: heavy AI infrastructure spending that has yet to show a clear payoff, and investor fears that AI agents from companies like Anthropic and OpenAI could weaken demand for its core software products.

Stock suffers: The company's stock is down 24% in the first quarter, on track for its biggest loss since its 27% drop in the fourth quarter of 2008.

It's by far the weakest performer among the Magnificent Seven tech giants at the start of the year, with an index tracking the group down 13% over the same period.

Long-term value? Some analysts say the stock now looks cheap, trading below 20 times next 12-month earnings. Others warn that software growth could slow and margins could come under pressure as customers turn directly to AI vendors.