The City That Publishes The Terms

Almost every startup programme on earth is sold in adjectives. This one publishes its parameters. An eight-week accelerator run by the Dubai Future Foundation puts an AI company inside a government services brief alongside more than 20 public entities, takes no equity, and pays for accommodation and travel. A second programme, six weeks long with Meta, ran design thinking and rapid prototyping on the open-source Llama 3 stack across energy, transportation, aviation, tourism and retail. A third, nine weeks under the Future Accelerators, pairs a company with a government entity on a challenge cycle. Duration, equity and who pays are the three fields that decide what a programme costs you, and all three sit on the page instead of in a call.

DIFC runs the largest dedicated AI cluster in the region: phase 1 holds 75 or more businesses against a stated target of 500 or more by 2028, with innovation and AI licences, coworking, offices and access to an NVIDIA DGX supercomputer, open year round. Silicon Oasis announced a second one in February 2026 at AED 5 billion and 2 million square feet, targeting 100 or more AI companies, five research labs tied to MIT, Stanford, MBZUAI and Carnegie Mellon, and 50,000 professionals trained by 2030, with the first phase occupied in the fourth quarter of 2026 and 20 incubator places in the first cohort. Same city, two doors, different terms. The licence and the desk come from one. The subsidy and the lab come from the other.

The Silicon Oasis package lists what it will pay for. Corporate tax at zero on a 50-year holiday. Rent cut by 50 per cent for three years. AED 100,000 of GPU compute credits for a startup. Research and development grants to AED 5 million. A licence inside 48 hours. Around those sit Dtec, a 12-month hub with a free zone setup and an investor network, and in5, run under TECOM Group, with coworking and incubation across tech, media and design. The itemisation is the useful part. A number that is written down is a number you can hold an authority to at renewal. A number described as generous is not a number at all.

Run the published comparison table and one column decides most of the shortlist. The government accelerator takes no equity. The Meta incubator states no equity and does not state travel. The DIFC cluster and the Silicon Oasis district are not applicable, because neither invests. Dtec and the Future Accelerators publish no equity take either. Read that column carefully. Free is not free money. It is a programme paid for by something other than your cap table, usually a government budget or a corporate strategy, and both of those change. Where the terms go quiet, the silence is the term. The probability that an unstated cost lands on you is not zero. Most applicants price it at zero and find out in week three.

A programme that publishes its terms has already done the comparison for you.

Status is a published field, and it separates an opening from a waiting list. The government accelerator closed cohort 2 in August 2025 and has announced no cohort 3. The Meta incubator launched in June 2024 and has run multiple cohorts since, with the next intake unannounced. The DIFC cluster and Dtec take applications year round. The Silicon Oasis intake is open now against a first phase in the fourth quarter of 2026. Intelak works aviation and tourism. TURN8 operates as venture capital as a service. Startupbootcamp runs a smart city track for city-facing AI. A deadline that has passed is information. An intake with no date is also information, and it is the kind most applicants skim past.

Three honest gaps sit in the published record. Travel is paid by one programme and unstated by two others, which means the flight and the hotel are yours until somebody confirms otherwise in writing. Cohort 3 of the accelerator carries no date, so the only route left is to monitor the foundation directly. And the Silicon Oasis numbers are announced targets rather than delivered ones: AED 5 billion, 2 million square feet, 100 companies and 50,000 trained describe what the announcement promises by 2030, not what stands today. That distinction is the whole discipline. A brochure tells you what a city wants you to feel. The published terms tell you what it will actually pay for, and almost nobody reads the terms first.