Double slashes often appear where two path segments are joined. One ends in /, and the next begins with it, producing an address such as /insights//checkout/.
Servers do not all handle that URL alike. Some redirect it, others serve the same content or return an error. Inspect the generated link and the actual response.
Narrow down the source
Fictional example: only article cards contain the doubled separator. The menu and sitemap use the correct paths. That points towards the card component or its data preparation rather than a site-wide server fault.
Compare the broken address with its intended single-slash version. Record whether both serve content directly and which canonical URL each page outputs. Then locate the code combining the base path and slug.
The sequence https:// requires two slashes. Characters outside the ordinary path may also have a different meaning. Replacing every double slash in a complete URL would damage correct addresses.
Fix path construction with a URL library and defined input rules. Update published internal links to the chosen address; an additional redirect can catch old requests. Check that its destination does not trigger another normalization rule and start a new round of redirects.
Related check: Check a canonical redirect destination.
For an identified path-construction fault, you can request a scoped change through Frontend optimisation.