Apple's home hub eve, TCS's growth problem, and the APK scam hitting Indian phones
Apple shows its smart home play tomorrow without a keynote, TCS keeps hiring while dollar growth crawls, a police warning on fake-bank APKs, a two-minute way to report scam numbers, and Nadella's AI emergency brake.
- Long Press · Daily Five12 Oct 202601 / 05
commentaryApple's home hub gets its moment tomorrow, and there's no keynote
Apple's smart home push gets its day on Tuesday, October 13, at 9am Eastern, which is 6:30pm in India. It isn't even a livestream, just a private 'experience' in New York with press releases and hands-on videos to follow. The rumoured star is HomeView, a screen-plus-speaker hub on an A18 chip, in two versions, with no App Store, shipping around mid-November. It's really a Siri test with a screen attached.
The problem
Apple has sold HomePods and Apple TVs for years without ever owning the centre of the home. Amazon and Google got there first with Echo Show and Nest Hub, and Siri has been the weak link the whole time.
What changed
Per MacRumors, Apple has confirmed only the date, the New York venue and the 'Welcome home' tagline. Bloomberg's Mark Gurman reports the hub is called HomeView, comes as a countertop model with a speaker base and a wall-mounted one, and ships around mid-November, with a new Apple TV 4K and HomePod mini arriving October 30, as MacRumors summarised.
The catch
Everything beyond the date is rumour, and there's still no price. A display with a limited set of Apple apps and no App Store only makes sense if the new Siri is genuinely good at running a house, and nobody outside Apple has used it for long yet.
What it means for you
Don't pre-order on launch-day vibes. Watch the first week of reviews for how Siri handles real commands, and remember Apple's home gear usually takes a while to reach India.
- Long Press · Daily Five12 Oct 202602 / 05
newsIndiaTCS keeps hiring. Its dollar revenue barely moved.
TCS's September-quarter numbers are a decent reality check on the 'AI will wipe out Indian IT' talk. Headcount rose by 4,258 to about 5.98 lakh, the third straight quarter of net additions, and the company puts its AI business at a $3.1 billion annual run rate. But dollar revenue grew just 2.4% from a year ago. If I worked there, that number would worry me more than any chatbot.
The problem
For two years the fear around Indian IT has been that AI tools would let clients do more with fewer engineers, and TCS's workforce shrank in FY26, which fed that story.
What changed
In results out on October 8, TCS reported Q2 FY27 revenue of Rs 73,188 crore, up 11.2% in rupees, and net profit of Rs 13,934 crore, according to Business Today. The company's own fact sheet shows dollar revenue of $7.64 billion, up 2.4% year on year and 2.8% in constant currency, with a $9.6 billion order book and attrition down to 13.3%.
The catch
Rupee growth flatters the picture because a weaker rupee inflates it. The AI run-rate figure is TCS's own and doesn't say how much of it is new money versus old projects relabelled.
What it means for you
If you're job-hunting in IT, hiring is steady rather than collapsing, but slow growth means raises and fresher pay stay tight. HCLTech reports today and Infosys on October 23, which will show whether this is TCS or the whole sector.
- Long Press · Daily Five12 Oct 202603 / 05
tipIndiaDon't just block the scam number. Report it on Chakshu.
Blocking a scam number protects you. Reporting it can protect the next person. In the Sanchar Saathi app or on sancharsaathi.gov.in, open Chakshu, pick whether it came as a call, SMS or WhatsApp, add the number, the time and a screenshot, and verify with an OTP. The Department of Telecommunications checks it and can disconnect the sender.
The problem
Most of us block a fake 'bank' caller and move on, which leaves the same number free to try the next hundred people.
What changed
Chakshu is DoT's reporting tool inside Sanchar Saathi for suspected fraud calls, SMS and WhatsApp messages, covering financial fraud, impersonation and fake offers. As Factly explains, DoT verifies each report and can re-verify the subscriber, cut the connection or refer it to police, and the reporter's identity stays confidential.
The catch
It's for attempts, not losses. If money has already left your account, Chakshu won't help you get it back, so call 1930 or file at cybercrime.gov.in straight away.
What it means for you
Next time a dodgy call or APK lands on WhatsApp, take the screenshot before you block, then spend two minutes on Chakshu.
SourcesFactly ↗Outlook Money ↗
- Long Press · Daily Five12 Oct 202604 / 05
tipIndiaYour bank will never send you an app on WhatsApp
Haryana's Rewari Police put out a warning on October 11 about .apk files sent over WhatsApp, Telegram and SMS by people posing as bank, courier or electricity staff. Install one and the malware sits in the background reading OTPs and opening your banking apps. The fix is boring and it works: keep 'Install unknown apps' off, install only from the Play Store, and never put AnyDesk on your phone because 'customer care' asked.
The problem
APK malware gets past the usual advice about not sharing OTPs, because once it's installed, nobody has to ask you for anything. It just reads them.
What changed
Rewari SP Hemendra Kumar Meena issued the advisory during Cyber Security Month, saying these cases are rising quickly, per The Tribune. The files arrive dressed up as a bank KYC update, a courier tracking app or an electricity bill, and give criminals remote access to messages, contacts, the gallery and banking apps.
The catch
Android warns you before sideloading, but people tap through warnings when the message looks urgent and the sender sounds official. Older parents are the usual target, which makes this a family conversation more than a settings change.
What it means for you
Check that 'Install unknown apps' is off for WhatsApp and your browser on every Android phone at home. If someone has already lost money, call 1930 and file at cybercrime.gov.in.
SourcesThe Tribune ↗
- Long Press · Daily Five12 Oct 202605 / 05
hot takeNadella wants an AI emergency brake. It should've shipped first.
Satya Nadella posted on Saturday that we should 'assume a model is compromised and contain it from the start', with an authorised human able to pause it mid-task. He called it an emergency brake. He's right, and tamper-proof logs plus a kill switch outside the model are basic engineering. What bugs me is the order. Agents are already inside Office and Windows, and the brake is still a post on X.
The problem
AI agents now take actions, not just answer questions, and the usual setup trusts the model to police itself, which falls apart the moment it's tricked by a poisoned document or simply gets things wrong.
What changed
In an X post on October 10, Nadella called for separating the model from the harness that runs it, moving safeguards outside the model, logging every meaningful action as 'tamper-proof human readable evidence', and letting an authorised person shut a model down mid-task, as TechCrunch reported.
The catch
A CEO post isn't a product spec. There's no date, no standard and no promise that Microsoft's own Copilot agents will meet this bar, and the industry has a habit of writing safety principles after the launch rather than before it.
What it means for you
If your company is rolling out AI agents, ask the boring questions now: who can stop it, where are the logs, and can the agent edit them. If nobody knows, that's your answer.
SourcesTechCrunch ↗
Taps
Quick hits. One line each, for when a long press is too much.
- IndiaBloomberg ↗
Jio's IPO price band reportedly set at Rs 1,065 to Rs 1,119 a share
Bloomberg says bids run Oct 21 to 23 with listing Oct 28, valuing Jio at up to Rs 10.3 trillion. Unconfirmed, and it could shift.
- IndiaKotak Neo ↗
HCLTech reports Q2 results today, Infosys follows on October 23
After TCS's 2.4% dollar growth, the thing to watch is whether big deal wins are finally turning into revenue.
- GlobalFinancial Times ↗
Nvidia is in talks to buy or invest more in Reflection AI
Per the Financial Times, it could be an acquihire, the deal shape big tech keeps using to sidestep antitrust reviews.
- GlobalFinancial Times ↗
Japan's cyber chief declares 'a state of emergency in cyber space'
More than 20 major Japanese companies have reported cyberattacks, per the Financial Times.
- GlobalWired ↗
Publishing staff say books are getting AI help without author consent
Employees at HarperCollins, Simon & Schuster and Hachette told Wired the use is growing quietly.