The Newsroom
London — How companies and people are putting AI to work
Working together
Forward-deployed now means learning together
Wipro's plan for 1,500 embedded AI engineers joins a wider movement toward teams that build with clients and leave them able to continue alone.
Common Intelligence · London · 27 August 2026
Wipro and Google Cloud announced on 27 August that they are preparing more than 1,500 forward-deployed engineers. OpenAI and AWS have launched comparable organisations this year. The common idea is simple: important AI systems are built with customers, not delivered to them.
An embedded engineer can see the difference between the documented process and the one people actually use. That matters because the exceptions, controls and responsibilities that keep an operation safe are often held by the team rather than by the software.
The strongest version of the model treats those employees as partners in the build. They select the real cases, explain why an apparently unusual decision was correct, test the first releases and help define when the system must stop and ask for review.
AWS makes the goal explicit: customer engineers should move from observers to co-builders to autonomous operators. The engagement is not successful if capability remains with the supplier or if the team must request every future change from outside.
This approach also makes adoption less abstract. People learn the system while solving their own work, not in a separate training programme. They can see which parts remove repetition and which decisions still need their judgment.
The growth of forward deployment is therefore more than a services trend. It is a recognition that technical capability and organisational capability must be built together if AI is going to remain useful after the first launch.
Wipro and Google Cloud, 27 August 2026
OpenAI launches the Deployment Company, 11 May 2026
AWS announces a $1 billion forward-deployed AI organisation, 17 June 2026
The client team moves from explaining the work to reviewing, improving and owning the system. — Common Intelligence
The Outcome
Value
Personal productivity is not yet company performance
McKinsey finds that eight in ten respondents feel more productive with AI, while reported enterprise EBIT impact remains unchanged from last year.
Common Intelligence · London · 25 August 2026
McKinsey's August survey contains two figures that belong together. Eighty per cent of respondents say AI improves their individual productivity, yet only 37 per cent report any enterprise-level EBIT contribution, almost unchanged from last year.
People are already finding ways to draft, analyse and decide faster. The company captures less of that benefit when each method remains personal, when systems are not connected and when nobody owns the change to the wider workflow.
The organisations reporting the strongest results redesign the work itself. They combine leadership attention with operational rigour, treat risk controls as part of the build and pursue growth or innovation alongside efficiency.
For a team, that means agreeing how individual techniques become shared practice without turning useful experimentation into bureaucracy. A repeatable workflow needs an owner, a set of real evaluation cases, clear permissions and a visible route for exceptions.
For management, it means choosing a measure that the people doing the work can influence and the finance team can verify. The purpose of the system is not to make everyone busier with AI; it is to change a business outcome without obscuring who remains accountable.
The opportunity is to convert the gains employees already feel into capability the company can keep. That conversion happens through shared workflows, not through a larger licence count.
McKinsey, The state of AI in 2026, 25 August 2026
A shared workflow turns one person's useful method into capability the company can keep. — Common Intelligence
The Handover
Execution
The new work needs a visible human boundary
OpenAI's August research shows agents moving into legal, sales and other business functions. The next design problem is responsibility, not access.
Common Intelligence · London · 12 August 2026
OpenAI describes enterprise AI moving from assistance to execution. Agents are no longer limited to answering questions; they can use tools, create files and complete longer pieces of work for review.
The use is spreading quickly outside engineering. OpenAI reports strong growth in enterprise Codex activity across legal, sales, recruiting and marketing, where decisions depend on policy, context and relationships as much as on speed.
That does not remove the human role. It makes the boundary more important. Someone must decide what context the system receives, which actions require approval, which evidence is sufficient and when uncertainty should become a refusal.
Those choices should be made with the people who already carry the responsibility. They know the difficult cases, understand the consequences of a false positive and can tell the difference between a shortcut and a broken control.
A good deployment gives the system repetitive work it can perform consistently and gives people better information at the point of judgment. It also makes the handoff visible enough that a reviewer can understand what happened and challenge it.
People plus AI is not a slogan for preserving the old process. It is a practical design rule: use machines for scale and repetition, keep accountable judgment with named people, and build the connection between them so neither side is guessing.
OpenAI, From assistance to execution, 12 August 2026
Every automated step meets a named point of review, refusal or accountable decision. — Common Intelligence
2 August
Transparency becomes part of the product
Article 50 of the EU AI Act now requires relevant providers and deployers to make certain AI interactions and outputs clear.
Source
20 July
Training remains the main response
The ONS says businesses most often build AI skills by training or retraining the people already in the organisation.
Source
30 June
Self-sufficiency is now an explicit goal
AWS says its forward-deployed projects should leave customers with trained champions, documentation and systems they can operate.
Source
