These overlap. Never add them up. One hour of work can draw on two or three skills at once.
The future of work deserves its own champions.
Show us what you've changed. We'll show the world.
Recognising organisations making work better with AI. Every month. Across three regions. With evidence, not hype.
Submit your AI success storyNo entry fee · Evidence required · Independent jury · Monthly deadline: the 20th
Know a team changing the way work gets done? Put them on the map.
- 01Work ReimaginedMaking work smarter and more meaningful, by handing the drudgery to AI.
- 02Human PotentialTurning technological change into human opportunity: reskilled, redeployed, elevated.
- 03Inclusive FutureOpening the future of work to more people, including women and people with disabilities.
- 04AI ChallengerSmall organisations. Extraordinary impact.
- 05AI for GoodUsing AI to make a difference beyond business.
The Board of the Future Award
Recognising the boards leading enterprise transformation in the AI age.
- OversightHow the board governs AI across the whole enterprise
- LeadershipThe strategic choices it made, and the people it brought with it
- OutcomesMeasurable results for performance, work and people
Every month an independent regional jury selects outstanding practices from India, the GCC and Africa. Winners are announced in the monthly heat-map issue, starting in November.
A feature in the issue and the video episode, the winner's emblem for your website, reports and hiring, a place in the AIWoW register, and a place on the shortlist for the annual awards.
One AIWoW juror each from India, the GCC and Africa, with judges from global business and consulting. Jurors step out of any decision involving their own organisation or clients.
What government could do to unlock the prize in India
₹30,270 crore a year on offer, 9% picked up. Tellers, clerks and branch staff hold 40% of it. Much of what holds it back is cost, skills, data and trust, and governments can move all four. Two options, each drawn from something another country has already done.
An AI sandbox for smaller lenders, run by the RBI
Smaller banks, NBFCs and co-operative banks lack the data, computing and skills to test AI safely; only 21% of surveyed institutions are building or using AI.
Shared computing, test data and supervisors on hand, so lenders can trial credit, KYC and fraud models before they reach customers.
Reserve Bank of India, through the Reserve Bank Innovation Hub, with the IndiaAI Mission.
The UK's Financial Conduct Authority opened a "Supercharged Sandbox" with NVIDIA in June 2025, giving any financial firm computing, data and regulatory support; testing began in October 2025. FCA, June 2025
Credit, KYC and fraud work: about ₹4,720 crore a year (16% of ₹30,270 crore).
A legal gateway for banks to share fraud signals with each other
Mule networks run across many banks, but each bank's AI sees only its own slice. MuleHunter.AI is live in 26 banks.
A clear legal basis, with safeguards, for banks, payment firms and NBFCs to share red flags on customers, so fraud models see the whole network.
Ministry of Finance and the RBI, run through the Indian Digital Payment Intelligence Corporation.
Singapore amended its financial services law and launched COSMIC in April 2024; six banks share information on customers showing several red flags. Rajah and Tann, 2024 TaxTMI, March 2026
KYC, fraud and compliance work: about ₹1,900 crore a year (6% of ₹30,270 crore).
| ID | Option | Where | Opened | Status |
|---|---|---|---|---|
| P-030-01 | Kiranas Go Digital drive | India | Issue 030 | Proposed |
| P-030-02 | Half-price, pre-approved AI tools | India | Issue 030 | Proposed |
| P-030-03 | Job-redesign grants tied to nationalisation | UAE, Saudi Arabia | Issue 030 | Proposed |
| P-030-04 | Shared test kit for shopper-facing AI | UAE, Saudi Arabia | Issue 030 | Proposed |
| P-030-05 | Consent-based sharing of mobile money records | Kenya, Nigeria | Issue 030 | Proposed |
| P-030-06 | Open language models for traders | Kenya, Nigeria | Issue 030 | Proposed |
| P-031-01 | RBI AI sandbox for smaller lenders | India | Issue 031 | Proposed |
| P-031-02 | Legal gateway for sharing fraud signals | India | Issue 031 | Proposed |
| P-031-03 | Shared platform to flag financial crime | UAE, Saudi Arabia | Issue 031 | Proposed |
| P-031-04 | Supervised generative AI sandbox | UAE, Saudi Arabia | Issue 031 | Proposed |
| P-031-05 | AI sandboxes at central banks | Kenya, Nigeria, South Africa | Issue 031 | Proposed |
| P-031-06 | Live, consent-based open banking | Nigeria | Issue 031 | Proposed |
Options for GCC and Africa are in their editions. These are options for discussion, not advice to any government. Precedents are described as published by their sources; pools are our estimates from this issue's figures. No one pays to appear in this section.
Where the unclaimed prize sits, and who is funding it
Most of the prize has not been picked up, so most of the market is still unserved. Two theses for investors, the disclosed funding we found, and dated calls we will score in public.
A voice assistant for banking agents and tellers
Banks' financial inclusion teams and the companies that run banking correspondent networks.
Counter and agent assistant, the top-ranked AI solution in this edition.
Low today. It needs voice tools in every Indian language and devices at the counter.
Banking BHASHINI, the RBI's language AI project, and policy option P-031-01.
A large bank or correspondent network piloting voice for its agents.
Model risk and AI governance tools for lenders
Banks and NBFCs that must keep an inventory of models, a kill switch and bias tests under the RBI's draft guidance.
Supervising and checking AI, the new skills in this edition, sold as software.
High once the RBI finalises its guidance; the draft closed for comment in July 2026.
RBI's draft model risk guidance and the Governor's call for a full inventory of AI models.
The final guidance and its compliance date.
Disclosed rounds for Indian AI built for finance that we could confirm are few and modest: Lumiq raised ₹50 crore in a Series B led by Bajaj Finserv Ventures in June 2026 (AF.net); InRisk Labs raised US$27m led by Bessemer and Northpoint in August 2026 for AI-led reinsurance (Inc42); Pibit.AI raised US$7m led by Stellaris for insurance underwriting, mostly for US clients (Inc42). Money is going to insurance and data; the counter is unfunded.
| ID | Call | Check by | Confidence | Status |
|---|---|---|---|---|
| I-030-01 | At least one disclosed round of US$5m or more for an AI tool built for Indian kirana stores. | 30 Jun 2027 | Moderate | Open |
| I-030-02 | At least one GCC retail software company launches an AI product sold to retailers it signed up for e-invoicing. | 31 Mar 2027 | Moderate-high | Open |
| I-030-03 | A disclosed round of US$10m or more for an African company offering AI bookkeeping or credit to informal traders. | 30 Jun 2027 | Moderate | Open |
| I-030-04 | A sovereign-linked fund outside Saudi Arabia leads a retail AI round of US$20m or more in the GCC. | 31 Dec 2027 | Low-moderate | Open |
| I-031-01 | A disclosed round of US$10m or more for an Indian company selling AI model governance or model risk tools to banks. | 30 Sep 2027 | Moderate | Open |
| I-031-02 | A Gulf central bank or financial regulator opens a supervised AI or generative AI sandbox for banks. | 30 Jun 2027 | Moderate | Open |
| I-031-03 | At least one bank in South Africa, Nigeria or Kenya announces it is licensing an outside AI credit engine. | 30 Jun 2027 | Moderate | Open |
Sector and market theses only. Nothing here rates, recommends or values any company or security, and none of it is investment advice. Companies are named only as examples of disclosed funding, from the public sources linked; some figures are not confirmed by the companies. No one pays to appear in this section, and it is separate from the AIWoW Awards jury.
Dated. Falsifiable. Scored in public.
Every forecast
Two minutes, two games: this week's AI word and AI crossword
A new word and a new crossword every Friday, Issue . Play here, then share your result without giving away the answers.
How we got these numbers
Every figure in this issue comes from public sources and our own analysis of how the work is done. Open any section to see the working.
AHow we sized the work
BThe new skills, and why banking needs more of someSupervising AI, checking its work, writing its instructions and looking after data.
CWhat we could not confirmRead this before you quote a number.
DSources
R Reported: a named report, filing or news story. A Our analysis of how the work is done. E Our estimate from the named sources.