Press & media
In the media, and how to work with us.
Everything the press needs is here, in one place.
Last reviewed: October 2026.
In the media
Entries appear after they have been checked: a named outlet/blog, a date, the role our principal played in the story, and a working link — plus an archived copy where one exists, because articles move.
-
January 2024 · The Quint: New Telecom Act: Are VPN Users and VPN Service Providers Under Threat?
[Principal], quoted in the story.
-
October 2023 · BBC: Inside India's deadly instant loan app scam - BBC News
[Principal], consultant in the story.
-
June 2022 · Medianama: How India can improve its cybersecurity directions #NAMA
[Principal], panelist in the story.
-
February 2021 · Vice: How Loan Apps Weaponise Your Data to Make You Pay
[Principal], quoted in the story.
-
January 2021 · Times of India: How Instant Loan Apps Became A Death Trap
[Principal], quoted in the story.
-
January 2021 · Times of India: Money-lending apps need stricter scrutiny: Experts
[Principal], quoted in the story.
-
June 2020 · Economic Times: India steps up vigil for cyber attacks from China after apps ban
[Principal], interviewed in the story.
-
February 2019 · New Indian Express: India's biggest diamond traders accidentally leaked customer data online Report
[Principal], quoted in the story.
-
February 2019 · Huffington Post: India's biggest diamond traders accidentally leaked customer data online Report
[Principal], quoted in the story.
-
January 2019 · Databreaches.net: California Department of Insurance Vulnerability Potentially Exposed Thousands of SSN and Other Personal Information
[Principal], quoted in the story.
-
January 2019 · Huffington Post: FoodPanda Breach Exposed Your Address, Number, And More
[Principal], quoted in the story.
-
October 2018 · Forbes: How Companies (And Hackers) Can Cash In On Smart Sex Toy User Data
[Principal], quoted in the story.
-
October 2018 · Coincrunch: Cryptojacking Alert Over 30000 routers, Countless computers Compromised in India. But there is a Fix!
[Principal], quoted in the story.
-
October 2018 · The Next Web: Crypto-jacking epidemic spreads to 30K routers across India
[Principal], quoted in the story.
-
August 2018 · Vice: Hackers Made Half a Million Dollars Pretending They Watched You Watch Porn
[Principal], quoted in the story.
-
April 2018 · Economic Times: CCTV cameras are watching you. But who's watching them?
[Principal], quoted in the story.
-
April 2018 · Krebs on Security: Transcription Service Leaked Medical Records
[Principal], quoted in the story.
-
February 2018 · Factordaily: India comes on the crosshairs of cryptojackers; even conglomerates targeted
[Principal], quoted in the story.
Speaking and media kit
Everything a journalist, editor or conference programmer needs to evaluate, book and credit a Banbreach principal. It also serves the workshop and referral channels: the same bios and topics are what partner teams ask for first.
Bios
Two lengths per principal are available: a 50-word version for programme notes and a 150-word version for features.
The principal, with verification policyHeadshots
Print-resolution headshots for each principal, will be supplied with a suggested credit line. We ask outlets to use the current photo rather than a cached one — the same principle as the corrections policy: the source of record lives here.
Topics we speak on
Each talk is grounded in a method we practise and publish — what it can establish, and what it cannot. We do not take speaking engagements on topics we cannot evidence, for the same reason we do not publish unverified credentials: the audience should be able to check what we claim.
-
Prompt injection as an untrusted-input problem
Why the injection is not the incident — and what security engineering's handling of hostile input transfers to agent systems, and what genuinely does not.
-
Governing AI agents: objectives, prohibitions, earned authority
The executive framing: what leadership actually controls when a workflow acts on its own, and what 'under control' means in testable terms.
-
What an AI workflow assessment should deliver
The dossier, not the demo: threat models, acceptance criteria, release gates and the economics behind a go / no-go decision.
-
Request versus grant: permissions for non-human actors
Why the agent's own plan is not permission, and how to log the gap so an audit trail exists before it is needed.
-
Forensic readiness: making AI agents investigable
What an agent should log so an incident can be reconstructed later — inputs, versions, tool calls, approvals — and what no log can establish.
-
Decision economics for AI workflows
Cost per accepted task, not vibes: how to make the operating cost of an agentic workflow visible and defensible in a budget review.
Formats: keynotes, executive working sessions, panels and podcasts.
Press contact route
Media and speaking enquiries go through the general form — category "Press or media enquiry" — and a principal replies personally, not an agency. Comment requests on AI security incidents are welcome; we will say clearly what we can and cannot speak to, and we do not speculate about live incidents we are not privy to. If you are on deadline, say so in the first line and we will treat it accordingly.
Contact the press route