MMOOZZGG.com

Glossary

What is an AI strategy

In short

An AI strategy is a company’s decision on why it needs artificial intelligence, which goals AI must reach, how much money and how many people are assigned to it, and in what order to proceed.

Business and adoption · AI strategy

MMOOZZGG.com editorsChecked: 1 min read

In depthAI implementation strategy: a quarterly plan with a price and metrics

For a child

Before the summer holidays, you decide what you want to learn: to swim, to ride a bike or to build a model airplane. You cannot do it all at once, so you choose the main thing, ask your dad for help and mark training days on the calendar. That summer plan is a strategy. Companies make a similar plan when they decide why they need smart programs.

In plain words

Without a plan, a company buys neural networks because they are in fashion: one department tries a chatbot, another draws pictures, and there is no common benefit. A strategy sets the direction. It records which goals the company wants to reach with AI, for example to cut the time it takes to answer clients or to lower the cost of paperwork. It also says who is responsible, how much money is assigned, which data may be used and what to do first. A working strategy fits on a few pages and is updated once a year.

For business

Per BCG data from 2024, companies with strong AI results take on about half as many directions as those that lag behind, and invest in a few priority ones. A strategy is needed to make this choice on purpose. It links business goals to specific processes and sets the budget, the team model (own staff, contractors, an outsourced leader), the approach to data and to 152-FZ, Russia’s personal data law, and the dependence on model suppliers. A common mistake is a strategy in the form of a slide deck without numbers or deadlines. A working version looks like a quarterly plan with a price and metrics, and it rests on a review of the company’s processes.

For a specialist

An AI strategy is an organization-level document that links business goals to a portfolio of AI use cases and the resources to deliver them. Typical sections: vision and goals, priority use cases ranked by effect and complexity, the operating model (centralized or federated), data and infrastructure (cloud, own servers, hybrid), risk management and compliance (AI governance), staff and training, and a roadmap with milestones. At the state level, Russia has the National Strategy for the Development of Artificial Intelligence until 2030, approved by Presidential Decree No. 490 of 10 October 2019, as amended by Decree No. 124 of 15 February 2024. A corporate document takes its priorities and data requirements into account.

Example

A logistics company approved an AI strategy with the goal of lowering the cost of processing one order. It chose the first processes, named an owner and banned sending client data to public services.

Related terms

Questions and answers

How is an AI strategy different from an implementation plan? The strategy answers the questions why and what comes first, for a year and beyond. The implementation plan sets out specific projects, deadlines, owners and the budget for the next quarter.
Who writes the AI strategy in a company? The CEO or the AI director is responsible for it, and it is prepared together with department heads, finance and IT. An outside consultant helps with a market review and an estimate of the effect.

Sources

Where companies use it

Not sure whether your company needs this?

AI solutions map
Contact us