Glossary
French: risque de confusion
The likelihood of confusion (risque de confusion) is the final condition for infringement where the goods and services and the signs are not identical — the French/EU counterpart of the Section 2(d) / DuPont analysis in US practice.
The likelihood of confusion is analyzed in view of the identity or similarity of the goods and services and the identity or similarity of the signs. The greater the similarity or identity on each axis, the easier confusion is to establish — the two factors interact, so strong similarity of signs can offset weaker similarity of goods, and vice versa.
Signs are compared on three levels — visual, phonetic and conceptual — through the eyes of the relevant public: the average consumer of the goods or services at issue, deemed reasonably well informed, observant and circumspect, though the level of attention expected of that consumer varies with what is being bought (a bottle of table wine invites less scrutiny than a piece of industrial equipment). This two-axis, interdependent comparison — sometimes referred to as the principle of interdependence — is what allows weak similarity on one axis to be compensated by strong similarity on the other. It has no single US label, but it echoes the way several DuPont factors are weighed together rather than in isolation.
The distinctiveness of the earlier mark also shapes the outcome: a highly distinctive or well-known earlier mark commands a broader scope of protection, making confusion easier to find, while a weak, largely descriptive earlier mark is protected only narrowly. The same idea drives US doctrine, where the strength of the senior mark is itself one of the core DuPont factors — and a genuinely famous mark can support a claim well beyond the DuPont analysis, through dilution.
Where both the sign and the goods or services are identical (“double identity”), the likelihood of confusion is presumed and need not be demonstrated — a structural shortcut with no exact US equivalent, where confusion must always be proven.
In French opposition proceedings, confusion is assessed on the registers: the signs and the goods and services as filed. The actual, practical conditions under which the goods are marketed are irrelevant. US counsel accustomed to arguing trade channels, price points, purchaser sophistication and marketplace realities should expect that evidence to carry no weight in an INPI opposition — the comparison is abstract, mark against mark, list against list.
This narrows the playing field considerably compared with a TTAB proceeding, where the full DuPont test invites evidence on trade channels, conditions of sale, actual confusion, and purchaser sophistication — all of which remain in play precisely because US practice does look past the bare register entries when the record supports it. Before the INPI or the EUIPO, none of that evidence is admissible: the only facts that count are the wording of the trademark registers and the goods and services lists as filed. The upside for brand owners is procedural simplicity and speed; the downside is that a sound business argument about diverging markets or sophisticated buyers has no home in a French or EU opposition.
Consider a US software company operating under the mark ACME CLOUD, challenged by an earlier French registration for ACME covering software services in Class 42. Before the INPI, the comparison stops at the register: the earlier owner need not show the two businesses actually compete, and the fact that one product targets enterprise IT departments while the other targets individual consumers is irrelevant. Before the TTAB, that same distinction in trade channels and purchaser sophistication could be argued as DuPont factors weighing against confusion — the same underlying dispute can diverge sharply depending on which side of the Atlantic it is litigated.
Likelihood of confusion is the central battleground whenever responding to a trademark opposition, whether the case is brought before the INPI or the EUIPO, and it resurfaces in any third-party rights validity challenge based on earlier marks. See also similarity of signs, similarity of goods and services, trademark imitation, and distinctive character, which drives how broadly an earlier mark is protected.