When she hears news from her cousin, the staff room, or the tea-stall, she has already heard it from JBIQ — and she knows what it means.
We deliver the news; we do not perform it. News is what is relevant today, not what is broadcast today.
Indians stitch together six broken news surfaces every day. WhatsApp for speed. YouTube for tone. The village paper for depth. Each one solves part of the job. None solves the whole job.
From the retired clerk on Aaj Tak to the auto-driver on Dhruv Rathee, every Indian news consumer makes a different broken trade. JBIQ composes the same two loops differently for each.
She is who the product is built for. If it works for Sunita, the rest follow.
Two coupled flywheels on one backend. The morning brief earns her trust. The verification keeps it. Without the morning, we are a fact-check tool nobody opens twice. Without the verification, we are a podcast.
The trust check. Sunita pastes a forward, holds the mic to a voice note, points her camera at a poster. JBIQ replies in 60 seconds in her language: what it is, whether it's true, what it means for her. Available on the JBIQ app and as a WhatsApp share-receiver.
News she can talk to. A personalised morning brief in her language, refreshed through the day, profile-aware and interactive any time. "Iske baare mein aur batao" turns the brief into an explainer. Same product, different read for every person.
A morning brief at 6:15. Breaking moments throttled to what actually moves her day. And, when a forward starts spreading, a verified version reaches her before the rumour does.
One push at 6:15. Maithili voice for Sunita, Hindi for Ramesh. Same product, different read. The brief is the open of the day. The rest of the surfaces earn a second open by being relevant, not loud.
A forward starts in one Parivaar group. Sixty more take it across districts. Our verification corpus catches the signal at 200K shares, classifies it, and fires the corrected version into the same chain — before Sunita has to choose what to believe.
Each door is gated by a different structural blocker. The aggregator stack, the platform shield, the surface presence. JBIQ is the only stack that walks through all three because we own the rails on each side: identity, editorial, and the WhatsApp loop.
Aggregators have the profile but not the editorial nerve. Publishers have the nerve but not the profile depth. Platforms have the loop but not the editorial mandate. Each door has a price the others won't pay.
Three different positions on the same value chain. JBIQ is the trust and personalisation layer above the others, not in competition with them.
"None of these can be merged. The rights stack makes the separation structural."
Three structural questions, three answers. Each opens to the version we'd give in a 90-second read.
JBIQ has three editorial jobs: we verify, we advise, we surface.
Verify means we call a claim verified, likely, unclear, or "we will not call this," with sources shown. Because we hold direct wire licences (PTI, Reuters, AFP), we grade publisher articles against canonical wire copy when verdicts conflict — a publisher-only stack could not do this.
Advise means on consumer and civic content that affects the user (health, money, schemes, civic process), we give practical guidance. Surface means we choose what goes into the personalised brief with attribution, and where content is partisan we surface both sides without picking one.
We do not author original news. Hard news appears as verified, attributed facts from three layers: government primary sources (gazettes, IMD, ECI, PIB, NDMA, scheme registries), wire services, and licensed publishers. Never from a JBIQ newsroom. The Editor-in-Chief owns standards, verdict process, and advisory quality — not a reporting desk.
Two options we considered and did not take: (a) Pure verifier, verification only, no advisory, no surfacing — rejected because verification alone cannot carry the daily ritual; without the personalised brief there is no flywheel between trust and habit, and the moat collapses to a single wheel. (b) Full newsroom, JBIQ authors hard news in the manner of Network18 or NDTV — rejected because authoring news puts JBIQ in direct competition with Network18 inside Reliance and with every licensed publisher outside it, reproducing the publisher-walks problem this strategy says aggregators cannot solve.
JBIQ holds three categories of news content rights, each licensed deliberately:
Display rights sit with JioNews; we ride on those existing publisher relationships. Real-time retrieval and citation rights are what JBIQ adds on top, signed bilaterally with 6 selective publishers: 4 Hindi-belt dailies (Jagran, Bhaskar, Hindustan, Amar Ujala) for district-level coverage, plus 2 English titles (Indian Express, Mint) for investigative depth. Training rights are signed selectively on 2 anchor publishers (1 Hindi, 1 English) where Iska Matlab's verifier benefits from corpus-level claim recognition. We do not buy training rights across all 12 publishers, because most publisher copy is wire-derived and we hold direct wire licences for that content.
All verdicts are owned by an independent editorial board with named external members. Network18 content is handled by external members only, under a mandatory internal-recusal rule. This pays publishers for what we use and gives external readers a structural answer to the conflict-of-interest read of a Reliance-owned product grading news.
Three approaches we considered and did not take: (a) Indian News Cooperative — co-found an industry body with 5-7 publishers for AI-era content standards. Rejected: leading newsrooms are politically factional and will not coexist as co-founders, and a Reliance co-founder reads as takeover rather than peer. (b) Network18-anchored — Reliance's own newsroom as primary editorial source. Rejected: concentrates editorial dependence on a single politically aligned publisher and destroys the independent-verifier brand on day one. Network18 is treated as one licensed publisher among several. (c) Full publisher upgrade across all 12, no wire deals. Rejected: ~70-80% of publisher content is wire copy, and the publisher cannot sublicense embedded wire content for AI use (ANI v. OpenAI, Delhi HC, 2024). The wires-first stack solves both layers cleanly.
Reliance has two news-related assets and they do different jobs. Network18 produces journalism — newsrooms reporting stories, channels, distribution and ad revenue. This is the publisher layer. JioNews aggregates third-party publishers and surfaces them inside the Jio device fleet. This is the aggregation layer. JBIQ verifies what is true, advises on what affects the user, and personalises what each user sees — drawing from government primary sources, wire services, and licensed publishers including Network18, with a verdict and advisory layer that neither of the other two carries. This is the trust and personalisation layer, sometimes described as the AI publisher layer.
The three roles also separate cleanly by content rights, which is why they cannot be merged. Network18 owns its content (in-house production, no external licence). JioNews holds display rights to redistribute publisher content verbatim with attribution; this is the aggregator-standard licence that does not permit AI use. JBIQ holds real-time retrieval and citation rights (with selective training rights on a few anchors), which permits paraphrase, synthesis across sources, and verdict issuance. None of these is permitted under an aggregator's display licence. The rights stack makes the three-way separation structural, not just functional.
JBIQ does not duplicate Network18 (we do not author news) and does not duplicate JioNews (we are not an aggregator). Network18 is treated as one licensed publisher among several, with the independent editorial board and Network18-recusal rule governing any verdict on its content. Network18 produces. JioNews aggregates. JBIQ verifies and personalises.