KPMG, one of the Big Four accounting firms, is cutting about 10% of its U.S. audit partners—roughly 100 senior figures—in a targeted restructuring of its audit business.

The move follows several years of unsuccessful efforts to encourage voluntary early retirements, leaving the firm with more partners than current demand justifies.

The firm says the cuts are not performance-related but part of a multi-year strategy to align the “size, shape and skills” of its audit leadership with evolving business needs.

Affected partners will receive financial exit packages and support in finding new roles.

KPMG’s U.S. audit practice has been growing and audits about 10% of public companies, but still trails major rivals.

The reduction reflects a broader industry correction after pandemic-era overhiring combined with unusually low attrition, which left firms with excess senior capacity.

Partner-level cuts are rare due to the cost and complexity of buying out equity stakes, making this move notable within the sector.

What is confirmed is a deliberate structural downsizing rather than a crisis response.

What remains unclear is whether the cuts will affect audit delivery or client retention as competitors seek to capitalize on any disruption.

Polymarket, a crypto-based prediction market where users trade on real-world outcomes, is at the center of a French investigation after suspicious bets on Paris temperatures coincided with abrupt, isolated spikes in data from a weather station at Charles de Gaulle Airport.

The case matters beyond a niche betting market because the disputed readings came from infrastructure tied not only to contract settlement but also to wider public meteorological systems.

What is confirmed is that Météo-France filed a police complaint after identifying anomalies in temperature readings on 6 April and 15 April.

Those spikes, reported as jumps of roughly four to five degrees Celsius within minutes before a rapid return to prior levels, did not align with surrounding stations, increasing suspicion that the measurements may have been artificially influenced rather than caused by ordinary weather variation.

The betting pattern is one reason the episode drew immediate scrutiny.

Traders placed low-probability positions on Paris reaching specific temperature thresholds and then collected unusually large profits when the anomalous readings pushed the contracts over the line.

In one widely cited instance, a wager of roughly thirty dollars generated close to fourteen thousand dollars.

On another date, a small position reportedly turned into a profit above twenty thousand dollars.

In practical terms, those returns matter less for their absolute size than for what they suggest: even modest manipulation of a single data point may be enough to produce outsized financial rewards.

The core mechanism is straightforward.

These weather markets settle against a designated external data source.

If that source records a temperature high enough to cross the contract threshold, the market resolves accordingly.

That creates a structural vulnerability: the platform may be technically secure while the real weak point sits outside it, in the physical or data-collection system used to define reality for settlement purposes.

Investigators are examining whether the sensor itself, or the immediate environment around it, could have been interfered with.

One theory under discussion is a highly localized heat source applied near the instrument, enough to alter the reading briefly without affecting nearby stations.

What remains unclear is whether any interference, if it occurred, was physical, digital, opportunistic or coordinated, and whether anyone with prior knowledge of the data pathway played a role.

The significance for French authorities goes well beyond gambling.

The same measurement network feeds aviation operations, forecasting and public weather services.

A brief distortion in one location does not mean the broader national system failed, but it does show how a public-data node can become financially attractive once a private market ties money directly to its output.

The episode also exposes a wider design issue in prediction markets.

These platforms promise price discovery based on crowd judgment, yet many contracts depend on a single authoritative source or a narrow chain of reporting.

That arrangement is efficient when the source is stable and hard to influence.

It becomes much riskier when the source is public, localized, physical or otherwise vulnerable to tampering.

The problem is not only insider trading in the classic sense of privileged information; it is the possibility of influencing the event or measurement itself.

Polymarket has already adjusted at least some of its Paris weather market rules by shifting away from the airport station as the reference point.

The company has also tightened language around manipulation and trading on outcomes a participant can influence.

Those steps may reduce immediate exposure, but they do not fully resolve the broader question raised by the French case: how a market built on external truth should function when that truth can be nudged, however briefly, by actors chasing profit.

The affair lands at a moment of growing scrutiny for event-based betting more broadly.

Regulators and policymakers have already been wrestling with questions about insider knowledge, market integrity and the ethics of wagering on politically or socially sensitive outcomes.

This case sharpens that debate because it is not only about who knew something first.

It is about whether the incentives created by such markets can encourage direct interference with the underlying facts used to settle them.

For now, the French inquiry remains focused on whether the temperature anomalies reflected tampering and, if so, who was responsible.

The answer will determine whether this becomes a narrow criminal case or a more consequential warning about a new category of risk: public data infrastructure turning into a target the moment a market starts paying for the right reading.