ISP throttling — deliberately slowing specific traffic types rather than a genuine congestion response — is harder to prove than it used to be, since most large ISPs have gotten more sophisticated about disguising it as network management rather than a policy targeting specific services. Here’s the current state of documented and suspected throttling cases, how to actually test for it yourself, and what a VPN does and doesn’t fix.
What Counts as Throttling vs. Ordinary Congestion
Not every slowdown is throttling. Genuine network congestion during peak hours (typically 7–11pm local time) affects all traffic on a shared node roughly proportionally, while true throttling is selective — specific protocols, ports, or destination services get slowed while everything else on the same connection runs at normal speed. The distinguishing test is whether a VPN, which encrypts and obscures the traffic type from the ISP’s deep packet inspection, restores full speed to the affected service while leaving everything else unchanged. If speed only normalizes for the previously-slow service after enabling a VPN, that’s a strong signal of selective throttling rather than general congestion.
Documented Historical Cases
The most well-documented throttling episodes remain the reference cases even years later: Comcast’s 2007 BitTorrent throttling (which led directly to the FCC’s original net neutrality enforcement action), AT’T’s throttling of unlimited-plan customers after they crossed a data threshold (settled with the FTC), and Verizon’s throttling of Netflix and YouTube traffic during a 2017 trial confirmed by researchers at Northeastern University and Queens College using their independent measurement tool. These cases matter less for their specifics now than as proof of concept — they establish that selective, protocol-aware throttling is technically real and has been deployed at scale by major carriers, not just a user-side conspiracy theory.
Current State: Mobile Carriers
Video-streaming throttling on mobile plans is now largely disclosed rather than hidden — most major US carriers explicitly cap streaming video resolution (commonly to 480p–720p) on mid-tier unlimited plans as a stated plan feature rather than covert throttling, and customers can usually pay for a higher tier to remove the cap. The more contentious current practice is network “deprioritization” during congestion for the cheapest unlimited tiers, which carriers disclose in the fine print but which functions identically to throttling from the user’s perspective — your data is technically “unlimited” but demoted to lowest priority whenever the local tower is busy.
Current State: Home Broadband
Overt protocol-based throttling on fixed broadband is less common in 2026 than a decade ago, partly due to regulatory scrutiny and partly because encrypted traffic (the overwhelming majority of internet traffic now runs through HTTPS/QUIC, unlike the unencrypted BitTorrent traffic Comcast throttled in 2007) makes deep packet inspection targeting far harder to execute reliably. The more common current complaint is applications-layer throttling by streaming and gaming services themselves adjusting bitrate based on detected congestion — which looks identical to ISP throttling from a speed test but originates from the service, not the network operator. This distinction matters for diagnosis and is frequently confused in user reports.
How to Test for Throttling Yourself
Run a baseline speed test to a generic server (not the throttled service) at the same time as testing the suspected-throttled service directly — a large gap where the general test is fast but the specific service is slow is the first signal. Then repeat the same service-specific test through a VPN; if speed to that specific service jumps back to baseline while the general speed test result stays the same, that strongly suggests selective throttling rather than congestion or server-side rate limiting. Tools built specifically for this — the Neutrality/Network Interference tools from Princeton’s CITP or the M-Lab-backed FCC Speed Test app — log results in a way that can support a complaint, unlike a casual speed test screenshot.
Comparison Table: Diagnosing the Cause
| Symptom | Likely Cause | VPN Fixes It? |
|---|---|---|
| All traffic slow at peak hours | Genuine network congestion | No — congestion affects VPN traffic too |
| One specific service slow, others fine, VPN restores speed | Protocol/service-based ISP throttling | Yes, typically |
| Video capped at low resolution on mobile plan | Disclosed plan-tier video cap | Sometimes, depends on carrier’s detection method |
| Streaming quality drops regardless of connection tested | Server-side adaptive bitrate (not ISP) | No — the throttling is happening at the streaming service, not your ISP |
What a VPN Does and Doesn’t Fix
A VPN genuinely defeats protocol/destination-based throttling because it encrypts your traffic and routes it through a single tunnel, hiding the underlying service from your ISP’s inspection tools. It does not fix general congestion (a VPN’s tunnel is still subject to the same last-mile bandwidth limits) and does not fix server-side adaptive bitrate throttling from the streaming service itself, since the slowdown originates past the VPN endpoint. Some ISPs have also started throttling known VPN server IP ranges specifically in regions with stricter internet controls, which is a separate and growing problem from traditional protocol throttling.
FAQ
Is throttling illegal?
It depends entirely on jurisdiction and current regulatory framework — net neutrality rules have been repealed and reinstated multiple times in the US at the federal level, and several states have passed their own net neutrality laws independent of federal status, so legality varies by where you live.
Does a VPN slow down my connection even without throttling present?
Yes, typically by some margin due to encryption overhead and the extra routing hop, though a well-run VPN with a nearby server often costs only 5–15% of raw speed — far less than what throttling itself costs if it’s actually occurring.
How do I file a complaint if I find real throttling?
In the US, the FCC accepts consumer broadband complaints directly, and documented, timestamped, VPN-comparison test results (not a single screenshot) make a materially stronger case.
Verdict
Real, selective ISP throttling is less common on fixed broadband today than a decade ago but still very real on the cheapest mobile plan tiers, mostly now disclosed as “deprioritization” rather than hidden. Before blaming your ISP, rule out server-side adaptive bitrate and genuine congestion with a same-time, service-specific VPN comparison test — if the VPN run alone restores full speed, you likely have a real throttling case worth documenting and, if it recurs, reporting.
