Large Staffing Firms Are Making Advances with AI

For a couple of years the big staffing groups spoke about artificial intelligence the way most companies do, in the future tense. It was coming, it was promising, it would eventually change the work. In their summer 2026 results, the tense changed. The numbers are no longer projections. They are operating metrics, reported to investors alongside revenue and margin, and they describe a recruiting process that has already been rebuilt around machines.

For a Canadian independent competing against these firms for the same clients and the same candidates, the shift is worth reading closely. This is no longer a question of whether the majors will use AI. It is a question of how far ahead they already are.

Adecco: seven agents and a million conversations

The most concrete disclosure came from the Adecco Group, which reported that it had surpassed 1.2 million AI-powered candidate interactions across ten countries, including 250,000 completed interviews spanning 50,000 jobs. The results it attached to those numbers are the ones that should focus a competitor's attention: a 50 percent reduction in time-to-deliver, fill rates above 80 percent, and candidate satisfaction scores averaging 4.3 out of 5.

Adecco now runs seven distinct AI agents across the recruitment lifecycle, covering pre-screening, talent-pool management, recruiter support, customer service, onboarding, and voice interaction. The most telling of them is the Redeployment Agent, which reconnects a candidate to a new opportunity within days of an assignment ending and builds a personalized profile of what to offer them next. That is the single hardest thing a staffing firm does, keeping good temporary workers engaged between placements, and Adecco has handed a meaningful share of it to software. Fully 51 percent of all these interactions happen outside traditional working hours, which is to say at the times a human recruiter is unreachable and a candidate is often actually free.

"At 1.2 million interactions, this is no longer about potential, it's about proven impact," chief executive Christophe Catoir said, framing the strategy as "talent, tech and touch," a deliberate signal that the human element is being repositioned rather than removed.

ManpowerGroup: the productivity dividend

ManpowerGroup's second-quarter results, reported July 23, told a similar story from the inside of a recovery. President Becky Frankiewicz reported a 67 percent decrease in time-to-fill over nine months of AI implementation, and said that 30 percent of candidate interviews now take place outside business hours through automation. The company is targeting $200 million in permanent cost savings by 2028 and recently sold its Jefferson Wells business in the United States for $88 million, a portfolio pruning that pairs naturally with a bet on doing more with less human overhead.

Crucially, this is happening as the business itself turns upward. The Manpower brand posted its fifth consecutive quarter of growth, with US revenue up 16 percent, and chief executive Jonas Prising said the brand had "moved from stabilization into a recovery globally." The AI investment, in other words, is not a defensive cost-cutting exercise carried out in a downturn. It is being layered onto rising volume, which is the more dangerous kind of efficiency for a competitor to face, because it compounds.

Randstad: efficiency in service of the flex thesis

Randstad, reporting the same week, framed its own recovery around a structural argument that explains why the AI investment pays off. Clients, chief executive Sander van 't Noordende said, "prefer to work with flexible workers because that obviously gives them the revenues, but not the risk." In an economy where employers want capacity without commitment, the firm that can source and deploy contingent workers fastest and cheapest wins the volume, and AI is the lever on both speed and cost. Randstad's North American business grew 4 percent in the quarter, with its US operations up 5 percent, the same regional strength visible across all three majors.

What a Canadian independent should take from this

Read together, the three sets of results describe an industrializing of recruiting that a smaller firm cannot match dollar for dollar, and should not try to. No independent Canadian agency is going to build seven proprietary AI agents or fund a $200 million transformation program. The instinct to compete on the majors' terms, by out-automating them, is a losing one.

The more useful lesson runs the other way. The majors have effectively published a map of which tasks now yield to automation: pre-screening, scheduling, first-round interviewing, redeployment outreach, and the after-hours candidate contact that used to fall through the cracks. That map is available to everyone, including the vendors who sell these capabilities to smaller firms as off-the-shelf tools. An independent does not need to build a Redeployment Agent to learn the lesson that redeployment outreach should be automated and constant rather than manual and sporadic. The capability is increasingly rentable.