AI in Iraqi companies: where it really saves cost in a hard year
Only a tenth of working-age Iraqis use AI tools. Where AI really saves cost, where it wastes time, and five steps for a small trial whose results you measure yourself.
In the AI Usage Index from Microsoft's AI Economy Institute, reported by Shafaq News in November 2025, Iraq ranked 86th in the world and among the lowest in the Arab region: only 10.3% of its working-age population uses AI tools (Shafaq News).
That sounds like bad news. It also means that an Iraqi company that starts well now is ahead of most of its competitors. In a hard year like 2026, the question is not "should we use AI?" but: where does it really save cost, and where does it waste time and money?
The numbers first
What global studies say
- Wide use, smaller impact: In McKinsey's 2025 survey, 88% of respondents said their organisations use AI regularly in at least one function. But only 39% saw an effect on operating profit at the level of the whole company, and most of them said that effect was below 5% (McKinsey).
- The difference is in how work is done, not the tool: The best-performing companies were about three times as likely to have redesigned their workflows fundamentally, rather than adding a tool on top of the old way.
- The biggest gain goes to new staff: In a study of 5,179 customer support agents, an AI assistant raised issues resolved per hour by about 14% on average, and by about 34% for new and less-skilled agents, with little effect on experienced ones (Brynjolfsson et al., NBER).
- Jobs change more often than they disappear: The International Labour Organization found that about a quarter of jobs worldwide have some exposure to generative AI, that clerical and administrative jobs are the most exposed, and that "transformation of jobs" is the most likely effect, not replacement (ILO).
And in Iraq
- Low government readiness: Iraq ranked 107th in the 2025 Government AI Readiness Index, with a score of 40.91 out of 100 (Shafaq News).
- Most people are online: Internet users reached 39.6 million, or 83.8% of the population, at the end of 2025 (DataReportal).
- The national strategy is still being prepared: Iraq announced its first national AI strategy in August 2024, and in April 2026 officials discussed building a national language model (Iraq Business News).
In short: the tool is available to you today, but the environment around you will not help much. The result depends on how you use it.
Where AI really saves cost
The simple rule: AI saves money where work repeats, where text or documents are the material, and where a person can check the result quickly.
1. Answering customers
Most customer questions repeat: price, availability, address, delivery date. An AI assistant can draft a first reply, and your employee checks it and sends it. This is where the customer-support study applies: the biggest gain goes to new staff, because the assistant hands them the experience of older colleagues.
2. Documents and correspondence
Proposals, first drafts of contracts, formal letters, monthly reports. Ask for a first draft, then edit it. The first draft is the slowest part of any writing, and that is what AI shortens.
3. Translation between Arabic and English
If you deal with foreign suppliers or partners, translating messages and offers every day takes hours. Current tools do this well, provided someone who understands both languages checks any important text before it goes out.
4. Data entry from invoices and documents
Moving numbers from a photographed invoice into a spreadsheet is dull and error-prone. Current tools can read the image and extract the data; your employee still checks the figures.
5. Quick analysis
You have six months of sales in a spreadsheet and no analyst? You can ask: which products fell? Which customers stopped ordering? The answer is a starting point for thinking, not a ready decision.
Where it wastes time and money
- Decisions that need context: the tool does not know your relationship with a customer or what the market did this week.
- Law and tax: a tool can give a confident, wrong answer. Any legal or tax point needs review by a specialist.
- Numbers without a source: ask for a statistic on the Iraqi market and the tool may invent one. Always ask for the source and check it yourself.
- Buying a big system before a small trial: this is the most expensive mistake. Start with one task.
What is specific to Iraq: four things not to ignore
- Iraqi dialect: the tools work better in Modern Standard Arabic than in dialect. Test a tool on real messages from your customers before you rely on it.
- Electricity and internet: Shafaq News lists chronic power cuts and weak networks among the main obstacles. Make the tool a helper, not a condition for work to continue.
- Customer data: do not put customer names, numbers and contracts into a public tool before you understand where that data is stored. Shafaq News also notes the absence of a clear legal and ethical framework for AI in Iraq, so for now the responsibility is yours. This point needs review by a legal specialist.
- Skills: a new employee learns faster when they use the tool alongside an experienced colleague who checks their work.
Five steps to start this month
- Choose one repetitive task: for example, answering price questions or writing proposals.
- Measure before you start: how many hours does it take each week, and how many errors happen?
- Run a thirty-day trial: one person, one tool, and a human check on every result.
- Write one page of usage rules: what may go into the tool, what may not, and who checks.
- Measure after the trial and decide: if it saved real time without new errors, extend it to a second task. If it did not, stop it and look for another task.
AI does not fix a bad process. If the process is chaotic, you get chaos faster.
What to watch
- The national AI strategy: and whether it will include rules on how companies use data.
- The national language model: discussed at the Ministry of Planning, and its effect on how well tools handle Iraqi dialect.
- New graduates: the University of Baghdad opened Iraq's first College of Artificial Intelligence for the 2025–2026 academic year, which means new skills in the labour market over the coming years.
In short
In a year when payments are late and costs rise, AI comes with big promises. But the global numbers say most companies use it and few see a real effect. The difference is not the tool, but choosing the right task, measuring, and rearranging the work around it.
Start small, measure honestly, and extend only what works.
Sources
- Shafaq News: Iraq last in the Arab AI utilisation index (Nov 2025)
- McKinsey: The state of AI in 2025
- Brynjolfsson, Li and Raymond: Generative AI at Work (NBER)
- ILO: Generative AI and jobs (May 2025)
- Shafaq News: Iraq and the AI divide (Aug 2026)
- Shafaq News: Iraq's race for AI (May 2025)
- Iraq Business News: AI strategy meeting at the Ministry of Planning (Apr 2026)
- DataReportal: Digital 2026 Iraq
Figures are reported as their sources published them, and some come from global surveys and indices that may not fully apply to the Iraqi market. This article is general information, not legal or technical advice for your situation. Use of customer data and any legal obligations need review by a specialist.
This article is general information, not legal, financial or tax advice for your situation.