AI and jobs have dominated the headlines for weeks. In the United States, employers now cite artificial intelligence as a leading reason for job cuts. In France, the official analysis stays far more cautious. For an SMB leader, the question is not which side to take, but what these numbers actually mean for a team.

In brief

• In the US, outplacement firm Challenger, Gray & Christmas recorded 87,714 AI-attributed job cuts between January and the end of May 2026, already more than the full 2025 total (54,836 cuts). In May alone, AI was cited in 40% of announced job cuts.
• France's Direction générale du Trésor (analysis published June 30, 2026) calls the net effect of AI on employment "uncertain": automation destroys some tasks, but productivity gains can also create labor demand elsewhere.
• According to the Observatoire des emplois menacés et émergents (Coface), up to 5 million jobs in France (about 16% of the workforce) hold positions exposed to AI automation by 2030. The study is careful to note this measures task exposure, not full job elimination.
• Amazon, Anthropic, Microsoft, and the OpenAI Foundation launched RAISE US in late June 2026, a worker retraining fund that has already raised over $500 million (target: $1 billion).
• For an SMB, the right response is to map exposed tasks before cutting roles, and to treat training as an investment rather than a budget line to cut first.

What the US numbers actually show

Challenger, Gray & Christmas, which has tracked stated reasons behind US layoff announcements since 2023, publishes a monthly breakdown of employer-cited causes. The 2026 trend is clear: in January, AI was cited in just 7% of job cuts. That share rose to 25% in March, 26% in April, and 40% in May, the highest monthly share on record.

Over the first five months of 2026, that adds up to 87,714 jobs whose elimination was justified, at least in part, by artificial intelligence. Technology accounts for most of these announcements, but finance shows up too, with 5,731 job cuts in May according to the same report.

One important caveat: Challenger, Gray & Christmas counts reasons stated by employers themselves. A company may cite AI for a reorganization whose real causes are actually mixed (economic slowdown, financial restructuring). Analysts commenting on these figures flag this as a known limitation.

Share of US job cuts citing AI as a reason, by month (source: Challenger, Gray & Christmas, June 2026 report).

And in France, what does the official analysis say?

France's Direction générale du Trésor, in an analysis published June 30, 2026, takes a noticeably more cautious tone than the US figures. The report notes that two effects pull in opposite directions: a substitution effect (automation replaces certain tasks) and a productivity effect (performance gains stimulate labor demand elsewhere in the economy). The Trésor concludes the overall net effect "remains uncertain," given limited historical data and the still-modest level of AI adoption among French companies.

The report does note that early AI-attributed job cuts in France concentrate, as in the US, in finance and IT, the most exposed sectors. It also highlights that 62% of French citizens view AI as a risk to employment (2025 Digital Barometer), a climate of distrust that leaders need to factor into internal communication.

Looking to 2030, the Observatoire des emplois menacés et émergents, backed by Coface, estimates that roughly 5 million French jobs hold positions exposed to AI automation, more than 16% of the workforce. The most exposed roles combine repetitive cognitive tasks: administrative, accounting, and legal positions, along with some engineering and analysis functions. By contrast, manual trades (crafts, repair, care work) remain, for now, largely spared. The study is explicit: it measures task exposure, not the wholesale elimination of jobs.

RAISE US: the major AI labs' response

Facing these rising numbers, Amazon, Microsoft, Anthropic, and the OpenAI Foundation announced on June 25, 2026 the launch of RAISE US, a nonpartisan worker retraining fund co-chaired by former US Commerce Secretary Gina Raimondo. The initiative has already raised over $500 million, targeting $1 billion, to fund pilot programs in four US states: wage insurance, incentives to retrain rather than lay off, AI-assisted career coaching, and short-term credentials.

This fund does not directly apply to French SMBs, but it sends a useful signal: the AI labs themselves recognize the need to plan for this transition rather than let it happen unmanaged. That same principle of anticipation should guide an SMB, at its own scale.

What should an SMB actually do?

Panic and denial are the two worst responses. Between them, a structured approach allows a company to plan ahead without overreacting to a single number.

FAQ

Will AI eliminate jobs at my company?

It mostly depends on how much of the roles involved consists of repetitive, non-relational tasks. US figures show a real acceleration in AI-cited job cuts, but France's official analysis (Direction générale du Trésor) still calls the economy-wide net effect uncertain.

Which jobs are most exposed to AI in France?

According to the Observatoire des emplois menacés et émergents (Coface), administrative, accounting, and legal roles, along with some engineering and analysis functions, concentrate most of the exposure. Manual and in-person service jobs remain less affected in the short term.

Should I cut headcount instead of investing in training?

The major AI labs themselves (Amazon, Microsoft, Anthropic, OpenAI) are now funding a retraining fund rather than letting layoffs be the only response. For an SMB, training existing staff usually costs less than hiring and retraining for different roles a year later.

Does RAISE US apply to French businesses?

Not directly. It is a US initiative currently limited to four pilot states. Its value for a French SMB is mostly as a signal: even the creators of AI tools consider retraining necessary, a useful argument for justifying a training budget internally.

Planning ahead instead of reacting

The 2026 numbers tell two different stories: a visible acceleration in the US, an acknowledged uncertainty in France. For an SMB, the best response stays the same either way: map, train, redeploy, relying on verifiable sources rather than media hype. To go further, explore our other AI resources for leaders and real company case studies that have already made this shift.