Lecturio, the medical education platform best known for its video course library and question bank, now markets an AI Tutor described as real-time Socratic guidance embedded in Q-bank sessions and grounded in its medical library (last checked August 2026). One product launch is a product launch; the pattern is the story. iatroX built its learning system around a Socratic Tutor; Lecturio has now reached for the same word and the same mechanism; Pastest has put AI inside the question with an assistance-shaped philosophy. When multiple companies independently converge on questioning-over-answering, the category is maturing, and this article takes Lecturio's launch seriously as evidence of exactly that, then asks the question maturity forces: once everyone claims Socratic, what actually differentiates a tutor?
What Lecturio has released
Per its current positioning: an AI tutor that operates inside question sessions rather than as a separate chatbot; that guides in real time while the learner works; that leans on graded support, hints and prompts before full explanation; and that grounds its teaching in Lecturio's own library, connecting struggling learners onward to the relevant course content. The design reads as a deliberate middle path between assistance and interrogation: keep the learner working, scaffold rather than solve, and use the platform's content depth as the safety net underneath the dialogue. For a company whose asset is a taught curriculum, attaching the tutor to the library is the architecturally honest move.
Why "Socratic" is suddenly everywhere
The word is appearing across medical AI for a defensible reason: the alternative failed pedagogically first. Answer-first AI made explanations free, and platforms watched learners consume fluent explanations without improving, because reception is not learning; the generation effect, learners improving by producing rather than receiving, is among the most robust findings in the science of learning, and questioning is how a system forces production. The precise ladder matters, though, and the word is doing a lot of work across products: an explanation tells; a hint narrows; a probing question makes the learner produce; and misconception diagnosis, the top rung, names the specific broken belief behind the production and fixes that. Marketing rarely distinguishes the rungs; buyers should, and the demo test is unchanged: answer wrongly, on purpose, and watch which rung the system reaches for.
Lecturio and iatroX, without manufactured conflict
The two implementations share the philosophy and differ in centre of gravity, which makes them easy to describe honestly. Lecturio's tutor sits inside a content ecosystem: its natural loop runs question, scaffolded dialogue, then onward into the video and library material that reteaches the topic properly, tutoring as the intelligent front door to a curriculum. iatroX's tutor sits inside a performance system: its loop runs error, interrogation to the named misconception, correction against cited clinical guidance, then spaced retesting and a revised plan, tutoring as the diagnostic engine of an adaptive loop. A learner who wants to be taught the topic again is better served by the first; a learner who wants the specific broken link found, fixed and verified is better served by the second; and a candidate could defensibly want both at different weeks of a campaign. Where Pastest fits, the assistance-shaped third philosophy, is mapped at /blog/pastest-lecturio-iatrox-qbanks-becoming-ai-tutors.
What still differentiates, once everyone is Socratic
The interesting competition moves up a level, to the system around the dialogue. Grounding: is the correction citable, to a library, or to clinical guidance itself? Learner history: does the tutor remember this learner's six weeks of errors, or only this question? Adaptivity: does the diagnosed misconception change which questions come next, or vanish into a transcript? Exam awareness: does the tutoring know which blueprint this learner faces, across which jurisdiction? Retention machinery: is the correction retested at intervals, or assumed to hold? Planning: does any of it alter what tomorrow contains? And continuity: does the phone session feed the same learner model as the desktop one? A Socratic dialogue with none of these is a good conversation; with all of them it is a learning system, and the gap between those two is where the next three years of this category will be decided.
What a genuinely intelligent tutor still cannot do
Honesty about the ceiling: no current tutor reliably models the learner's reasoning process directly, all infer it from answers and dialogue; none carries the stakes and noise of real clinical encounters, which is simulation's territory and ultimately the wards'; and none should author the learner's reflection or substitute for supervised practice. The claim maturity supports is narrower and still large: the questioning tutor, embedded in the question, with memory and a plan, is becoming the default expectation of serious revision software, and Lecturio's launch is one more company voting that way with its roadmap.
Frequently asked questions
Is Lecturio's tutor available on its existing plans?
Packaging and availability vary by plan and region and change quickly; check Lecturio's current pages rather than any article, this one included (last checked August 2026).
Does a video-course platform need a Socratic tutor at all?
Arguably more than anyone: passive video is the format most exposed by the generation effect, and a questioning layer is the correction. Attaching it to the library turns consumption into curriculum, which is the strategically sound version of the move.
How should a candidate compare the Socratic claims?
One protocol, run identically: three deliberate wrong answers on each platform, noting which rung, explanation, hint, question, misconception, the system reaches, and what changes afterwards. Twenty minutes, and more informative than every marketing page combined.