This isn’t a Make or AI issue. That RSS feed is now returning a 403 to automated requests, so Finextra is most likely blocking non-browser clients.
That’s why Zapier can still access it (they use different headers or are already whitelisted), while Make and AI tools get blocked. A VPN probably won’t help here since this is about request headers, not IP.
If you want to try something, use the HTTP module with a custom User-Agent header that looks like a browser. Otherwise, you’ll need a proxy or an official API if Finextra provides one.
So yeah, nothing broke on your side. The feed provider changed their rules.
thanks for your quick answer, point of rss is to be parsed by bots.. you know, and for those sites more they publish their news more they get traffic. so I think it is a cloudflare issue. but yes an idea is http request, I will look I’m not specialist with this
Totally agree with you. RSS is meant to be consumed by bots.
That said, many sites now put RSS behind Cloudflare or similar protection, and it ends up blocking anything that doesn’t look like a regular browser. So even though the feed is public, automated clients still get a 403.
Trying it via the HTTP module with a browser-like User-Agent is usually the easiest next step, and you don’t need to be an HTTP expert to test that. If that works, you’re good. If not, then it’s definitely Cloudflare being strict.
So yeah, your reasoning makes sense — this is more about protection layers than RSS itself.
I just tried an httpd get with User-Agent : Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/143.0.0.0 Safari/537.36 , which exactly mine, same. I guess it is by IP, the IPs of make are making too many requests everywhere
If a real browser User-Agent still gets a 403 from Make, then it’s almost certainly IP-based blocking (Cloudflare reputation or rate limiting). Make’s shared IPs get hit hard, so some providers just block them outright.
At that point there isn’t much you can do inside Make itself. A VPN won’t help either, since Make doesn’t run through your local IP.
Your realistic options are to use a proxy / scraping service with rotating IPs, or ask Finextra to whitelist Make IP ranges (unlikely, but possible). Otherwise you’ll need an official API or a third-party feed mirror.
So your conclusion is right. Nothing wrong with RSS or your setup, it’s IP reputation blocking.
That’s actually a solid approach. Using multiple RSS sources reduces dependency on a single provider and avoids getting blocked, and working from your own DB gives you full control and stability.
Once the data is in your DB, you can normalize, dedupe, enrich, and process it however you want without worrying about external rate limits or IP blocks.
Probably the most reliable long-term solution in this situation.
yes but it is big job because different scenarii are taking different type of data and it was for each of them an AI module choosing to treat the news.. then it needs to be flagged as used bu 1 scenerio and not the other, really complicated
Relying on multiple RSS feeds and processing everything from your own database reduces dependency on any single provider and avoids these access issues. With the way Cloudflare blocking works today, it’s often the most stable approach.
Thanks for sharing your findings. Useful for others running into the same issue.
What is your goal?
What is the problem & what have you tried?