Ivan Jackson
CEO & Founder
Ivan Jackson is the founder and CEO of WriteHuman, an AI software company focused on improving and refining AI-generated writing. He builds tools that help people turn robotic AI text into natural, human-sounding content. Based in Arlington, Virginia, Ivan works at the intersection of product, AI, and growth, building software used by millions of writers and professionals around the world.
78 posts
AIHumanize.io Review: Does It Beat AI Detectors?
AIHumanize.io is a functioning, budget-priced humanizer that passes detectors a solid majority of the time, but it finished 10th of 13 in HumanizerBench's July 2026 cycle and carries real billing, support, and account-lockout complaints. Fine for casual, price-sensitive writers; underpowered for anyone whose output has to hold up.
Grammarly Review 2026: Is Its AI Humanizer Worth It?
Grammarly is an excellent general-purpose writing assistant, but its AI Humanizer is built for clarity, not for changing AI-detection outcomes, and the benchmark data reflects that split.
Humbot AI Review: Does It Hold Up Against the Detectors in 2026?
Humbot is a mid-pack humanizer (HumanizerBench #5 of 13, 66.42 composite) that does fine against lenient detectors but clears GPTZero only 12.1% of the time, so it is a weak choice for anything that faces strict checkers.
GPTHuman AI Review: Does It Read as Human?
GPTHuman is a cheap, easy all-in-one AI humanizer that suits short, casual writing but frays on long-form work, with opaque ownership, no refund policy, and a headline success claim real users do not consistently reproduce.
![The [ud] Undetectable AI logo above the headline 'Is Undetectable AI Legit?' with the subhead 'a 2026 review of whether the writing holds up' on a light gray background.](https://www.writehuman.veo-tp.com/sb/storage/v1/object/public/blog-images/blog/1785764773690-ueqatz.webp)
Is Undetectable AI Legit? A 2026 Review of Whether the Writing Holds Up
Undetectable AI is a legitimate, easy-to-use humanizer with strong detector pass rates, but independent benchmark data shows it takes far heavier quality penalties than the top-ranked tool, so long-form writers pay for their scores in prose quality.
Claude's Watermark Punishes the Wrong People
Anthropic's new text watermark is easy to remove and easy to miss. The writers most likely to get flagged are the ones who did nothing wrong.
Claude Is Shipping an AI Text Detection API You Can Use Yourself
Anthropic confirmed it is shipping a text detection API that anyone can call to check for Claude's invisible watermark. Here is what that detector can actually establish, the limits Anthropic has published itself, and how WriteHuman is preparing for it.
Claude Is Now Adding Invisible Watermarks to AI-Generated Text
Anthropic is adding invisible watermarks to text generated by new Claude models. Here’s how the system works, what Anthropic says about its limits, and why users who don’t want to publish raw Claude output may choose to substantially rewrite it with tools like WriteHuman.
Introducing the WriteHuman Chrome Extension: Humanizing and Grammar in Every Text Box
The WriteHuman Chrome extension is live. It checks spelling and grammar in any text box with no account needed, and rewrites AI drafts in place with a human score, without a second tab or a single copy-paste.
Twixify Review: Breaking Down the Echowriting Approach
Twixify is a cheap, web-only AI humanizer whose echowriting voice-mimicry is its best idea, but inconsistent output, detector-dependent results, and recurring billing and cancellation complaints make it a gamble for anyone who needs reliability.
WriteHuman Takes #1 in the August 2026 AI Humanizer Rankings
WriteHuman finished #1 of 12 in the August 2026 HumanizerBench cycle with a composite of 76.69 and the highest detector pass rate in the field at 89.1% — all on the entry-level Basic plan.
BypassGPT AI Review: Where Your Text Goes and Whether It's Worth Paying For
BypassGPT is a legitimate, support-friendly AI humanizer with cheap annual pricing, but its operator is undisclosed, its refund terms are restrictive, and independent tests question whether its output reliably reads as human.