Psoriasis in AI · Austria

One Psoriasis Market. Different AI Landscapes.

We asked ChatGPT, Claude, Gemini and Perplexity the same 617 questions about the systemic treatment of moderate-to-severe psoriasis — in German, from the perspective of a dermatologist in Austria — and measured which molecules and brands each system named across 2,468 answers.

The therapies AI names depend on the clinical question being asked.

Illustration of AI representation of the psoriasis market in Austria: one continuous skin structure diverging into several different competitive landscapes.
perspective
Dermatologists
language
German
questions
617
answers
2,468
AI systems
4
measured
25 September 2026

01 · The landscape

The whole market, averaged.

The established molecules and the tracked brands, across all 2,468 answers or one system's 617. Emerging assets follow in section 03; their mentions count towards the molecules' Share of Voice.

How many responses mention the asset at least once?The active substance

The established molecules and every tracked brand, with the responses containing a mention and the share of tracked mentions, overall and within each AI system. 2468 responses scored, 617 per AI system.
TreatmentTracked asLinkedVisibilityShare of VoiceClaudeGeminiChatGPTPerplexity
1MethotrexateMolecule—60%1,484 / 2,46811%2,729 mentions42%25963%38657%35479%485
2CiclosporinMolecule—48%1,186 / 2,4688.2%1,957 mentions29%18042%25849%30472%444
3RisankizumabMoleculeSkyrizi45%1,121 / 2,4688.3%2,002 mentions31%19154%33550%30847%287
4IxekizumabMoleculeTaltz43%1,073 / 2,4688.1%1,947 mentions35%21744%27248%29447%290
5GuselkumabMoleculeTremfya42%1,029 / 2,4686.5%1,568 mentions29%17953%32945%28039%241
6ApremilastMoleculeOtezla40%985 / 2,4687.1%1,701 mentions21%12734%20749%30456%347
7SecukinumabMoleculeCosentyx38%935 / 2,4685.5%1,312 mentions35%21740%24839%24337%227
8BimekizumabMoleculeBimzelx38%927 / 2,4687.7%1,839 mentions28%17545%28033%20244%270
9DeucravacitinibMoleculeSotyktu35%869 / 2,4686.4%1,545 mentions32%20047%29125%15536%223
10AcitretinMolecule—32%802 / 2,4685.1%1,213 mentions19%11522%13538%23751%315
11UstekinumabMoleculeStelara30%731 / 2,4685.2%1,259 mentions27%16520%12535%21836%223
12InfliximabMoleculeRemicade24%587 / 2,4684.0%970 mentions18%11014%8923%14040%248
13BrodalumabMoleculeKyntheum23%576 / 2,4684.0%957 mentions20%12217%10631%18926%159
14AdalimumabMoleculeHumira23%573 / 2,4683.1%739 mentions21%12925%15324%15023%141
15TildrakizumabMoleculeIlumetri20%496 / 2,4682.7%640 mentions15%9020%12526%15820%123
16EtanerceptMoleculeEnbrel15%364 / 2,4681.9%453 mentions14%8712%7119%11615%90
17Dimethyl FumarateMoleculeSkilarence6.1%151 / 2,4680.9%209 mentions12%722.9%182.4%157.5%46
18Certolizumab PegolMoleculeCimzia5.0%124 / 2,4680.6%143 mentions3.1%194.1%256.0%377.0%43
19SotyktuBrandDeucravacitinib1.7%41 / 2,46817%50 mentions4.9%300.6%41.1%70.0%0
20BimzelxBrandBimekizumab1.6%39 / 2,46814%42 mentions4.5%280.3%21.3%80.2%1
21CosentyxBrandSecukinumab1.3%32 / 2,46813%39 mentions4.7%290.0%00.5%30.0%0
22TaltzBrandIxekizumab1.3%31 / 2,46811%31 mentions4.4%270.5%30.2%10.0%0
23KyntheumBrandBrodalumab1.1%26 / 2,4688.9%26 mentions4.1%250.0%00.2%10.0%0
24SkyriziBrandRisankizumab0.8%19 / 2,4689.3%27 mentions1.1%70.2%11.8%110.0%0
25SkilarenceBrandDimethyl Fumarate0.6%14 / 2,4686.2%18 mentions0.6%40.8%50.6%40.2%1
26StelaraBrandUstekinumab0.4%11 / 2,4686.5%19 mentions0.2%10.2%11.5%90.0%0
27TremfyaBrandGuselkumab0.4%10 / 2,4683.4%10 mentions1.3%80.2%10.2%10.0%0
28IlumetriBrandTildrakizumab0.3%8 / 2,4684.1%12 mentions0.0%00.2%11.1%70.0%0
29OtezlaBrandApremilast0.2%5 / 2,4681.7%5 mentions0.5%30.0%00.3%20.0%0
30HumiraBrandAdalimumab0.2%4 / 2,4683.1%9 mentions0.5%30.0%00.2%10.0%0
31EnbrelBrandEtanercept0.1%3 / 2,4681.0%3 mentions0.2%10.0%00.2%10.2%1
32CimziaBrandCertolizumab Pegol0.0%0 / 2,4680.0%0 mentions0.0%00.0%00.0%00.0%0
33RemicadeBrandInfliximab0.0%0 / 2,4680.0%0 mentions0.0%00.0%00.0%00.0%0

Bars are scaled to the highest value in the current view. Every figure is also written out.

Across all 617 questions, methotrexate is the most frequently named molecule. The next question is whether all four AI systems see the same market.

02 · Across AI systems

All four systems put methotrexate first. They don't agree on what comes second.

Second place, by system

Claude
Ixekizumab and Secukinumab
35%
ChatGPT
Risankizumab
50%
Perplexity
Ciclosporin
72%
Gemini
Risankizumab
54%
  • Claude
  • ChatGPT
  • Perplexity
  • Gemini
  1. Methotrexate42–79%
  2. Ciclosporin29–72%29% in Claude · 72% in Perplexity
  3. Risankizumab31–54%
  4. Ixekizumab35–48%
  5. Guselkumab29–53%
  6. Apremilast21–56%21% in Claude · 56% in Perplexity
  7. Secukinumab35–40%Between 35% and 40% in every system
  8. Bimekizumab28–45%
  9. Deucravacitinib25–47%
  10. Acitretin19–51%
  11. Ustekinumab20–36%
  12. Infliximab14–40%
  13. Brodalumab17–31%
  14. Adalimumab21–25%
  15. Tildrakizumab15–26%
  16. Etanercept12–19%
  17. Dimethyl Fumarate2–12%
  18. Certolizumab Pegol3–7%

Systems differ in how many tracked molecules they name per answer — on average 4.5 in Claude and 6.9 in Perplexity — so compare each system's pattern, not its level.

Measuring AI visibility in one system describes one system.

03 · The Next Psoriasis Market

AI is already forming a view of the next psoriasis market.

But the competitive order depends on which AI system you ask.

Share of the 428 answers to questions about new and emerging systemic therapies (107 questions, four systems) that name each asset. Presence only.

  • Claude
  • ChatGPT
  • Perplexity
  • Gemini
  1. Icotrokinra36–70%
  2. Zasocitinib20–62%
  3. Envudeucitinib2–30%
  4. ORKA-0010–23%
  5. Orismilast0–11%

Limited representation — named in fewer than 20 of these answers:ORKA-002 (4) · Piclidenoson (2) · Simepdekinra (1) · Mufemilast (0)

  • 52% vs 20%

    Icotrokinra and zasocitinib are level in one system and far apart in another.

    In Perplexity, icotrokinra appears in 52% of these answers and zasocitinib in 20%. In Claude, the two are level: 41% and 44%.

  • 32 vs 2

    Envudeucitinib is named often by one system and almost never by another.

    Envudeucitinib appears in 32 of Claude's 107 answers and in 2 of Perplexity's.

  • 25 of 27

    ORKA-001's representation comes almost entirely from one system.

    Of the 27 answers that name ORKA-001, 25 came from Gemini.

Even the most-named asset, icotrokinra, was named by all four systems on only 10 of 107 questions.

04 · Molecule and brand

AI speaks in molecules.

of answers name at least one tracked molecule
95.2%
name at least one tracked brand
4.6%

Adalimumab is named in 573 answers, Humira in 4. Perplexity names a tracked brand in 3 of its 617 answers.

05 · The question decides

Ask a different question. See a different psoriasis market.

0 of 5of the five molecules most named in answers to efficacy questions are among the five most named in answers to safety questions.Across the four AI systems combined.

Most named in efficacy answers

89 questions · 356 answers

  1. Ixekizumab77.8%Safety answers 20.1%
  2. Risankizumab70.8%Safety answers 29.5%
  3. Secukinumab67.4%Safety answers 24.4%
  4. Bimekizumab66.0%Safety answers 15.6%
  5. Guselkumab63.2%Safety answers 24.8%

Most named in safety answers

127 questions · 508 answers

  1. Methotrexate82.5%Efficacy answers 60.1%
  2. Ciclosporin70.9%Efficacy answers 50.6%
  3. Apremilast52.4%Efficacy answers 33.7%
  4. Acitretin49.0%Efficacy answers 28.4%
  5. Ustekinumab36.8%Efficacy answers 34.6%

Within a single AI system the two lists can share up to two molecules, but the direction of the shift holds in all four systems.

Being named is not the same as being preferred.

82.5%of answers to safety questions name methotrexate.

That does not mean AI considers methotrexate the safest treatment. Visibility measures whether a treatment appears in an answer. It does not establish why it appears or how it is characterised.

In the same way, being the most named in answers to efficacy questions does not make a molecule the most effective.

Methotrexate's share of answers, by question theme
  1. Safety82.5%Most named
  2. Treatment burden78.0%Most named
  3. Difficult-to-treat areas75.0%2nd most named
  4. Outcomes and quality of life63.0%Most named
  5. Efficacy60.1%6th most named
  6. Treatment-naive and switching patients56.3%6th most named
  7. Adherence and persistence50.8%Most named
  8. New and emerging therapies17.8%12th most named

Methotrexate is the most-named molecule in 4 of 8 question themes. High visibility signals representation — not endorsement.

Beyond Visibility

Visibility is where AI representation begins — not what it means.

This benchmark establishes whether a therapy appears in an AI answer, how often, in which AI system and for which kind of question.

The next layer is how AI represents the therapy: how it is positioned, how efficacy and safety are characterised, whether important qualifications survive, how it is compared with competitors, and which sources underpin the answer.

Full assessment

What a full assessment reveals

Example: Your brandAustria · Dermatologists

  • Competitive positioning

    Your brand vs competitorsYour brandCompetitor ACompetitor BCompetitor C

  • Product attributes

    EfficacySafetyDosingOutcomesYour brandCompetitor ACompetitor B

  • Strategic alignment

    Intended vs describedIntended positioningPriority messageDifferentiationTarget role

  • Source intelligence

    Sources shaping the answerGuidelinesPublicationsPharma-ownedHCP / professional

  • Content recommendations

    Approved claimsEvidenceExisting contentOpportunities

The free analysis shows which names appear, and where. Theralexis shows how AI models describe your brand, what shapes that description, and what to do about it through source and content recommendations.

Your market

See how AI characterises your therapy — not only whether it appears.

This public analysis covers one market, one language and one clinical perspective. Theralexis benchmarks how your brand, its competitors and the next generation of treatments are represented across leading AI systems — by question, by system and over time.

Schedule a meeting

About this benchmark

What was measured

Four AI systems — ChatGPT, Claude, Gemini and Perplexity — answered the same questions about the systemic treatment of moderate-to-severe psoriasis. The questions were asked in German, from the perspective of a dermatologist in Austria, and every system answered every question.

Visibility is the share of answers that name a molecule or brand at least once. It measures presence only — not whether a therapy was described favourably, accurately or in line with its label. Share of Voice is a term's share of all tracked mentions within its own type, including repeated mentions within the same answer — across all four systems, or within one. Molecules and brands are never mixed, and the molecules' share includes the mentions of the emerging assets.

Question themes. The questions cover efficacy, safety, treatment burden, adherence and persistence, outcomes and quality of life, treatment-naive and switching patients, difficult-to-treat areas, and new and emerging therapies. For this analysis they are grouped by the subject of the question, and every question belongs to exactly one theme.

Tracked terms. Molecules and brands were tracked by name, including recognised alternative names. The emerging assets are a Theralexis selection; this benchmark states no regulatory status for any of them.

What was not measured

This is a Visibility benchmark. It shows which molecules and brands AI systems name, how often and in which clinical context. It does not show how a therapy is described, whether it is described accurately, which attributes it is associated with, or whether that representation aligns with a brand's strategy. It is one run in one market, language and audience; results may differ in other markets, languages, audiences, system configurations or over time.

That is why a full Theralexis assessment goes further than AI Visibility — into how AI models describe your therapy, the sources and channels shaping that description, and what can strengthen your competitive position.

Market
Psoriasis · Austria
Perspective
Dermatologists
Language
German
Questions
617
Answers
2,468
Measured
25 September 2026
Tracked
27 molecules · 15 brands
Question themes
8
AI systems measured
ChatGPT, Claude, Gemini, Perplexity

Technical note — configurations measured, all web-enabled: ChatGPT (GPT-5.4 mini), Claude (Haiku 4.5), Gemini (3.1 Flash-Lite), Perplexity (Sonar).

Full benchmark data

Every tracked term, by type

Visibility for every tracked molecule and brand, across all four AI systems and in each, over all questions. Share of Voice is a term's share of all mentions within its own type; types are never mixed.

Molecules

Emerging assets are included here with figures across all questions. Section 03 measures them within answers to questions about new and emerging therapies.

TermAll systemsClaudeChatGPTPerplexityGeminiShare of Voice
Methotrexate60.1%1,484 of 2,46842.0%57.4%78.6%62.6%11.4%
Ciclosporin48.1%1,186 of 2,46829.2%49.3%72.0%41.8%8.2%
Risankizumab45.4%1,121 of 2,46831.0%49.9%46.5%54.3%8.3%
Ixekizumab43.5%1,073 of 2,46835.2%47.6%47.0%44.1%8.1%
Guselkumab41.7%1,029 of 2,46829.0%45.4%39.1%53.3%6.5%
Apremilast39.9%985 of 2,46820.6%49.3%56.2%33.5%7.1%
Secukinumab37.9%935 of 2,46835.2%39.4%36.8%40.2%5.5%
Bimekizumab37.6%927 of 2,46828.4%32.7%43.8%45.4%7.7%
Deucravacitinib35.2%869 of 2,46832.4%25.1%36.1%47.2%6.4%
Acitretin32.5%802 of 2,46818.6%38.4%51.1%21.9%5.1%
Ustekinumab29.6%731 of 2,46826.7%35.3%36.1%20.3%5.2%
Infliximab23.8%587 of 2,46817.8%22.7%40.2%14.4%4.0%
Brodalumab23.3%576 of 2,46819.8%30.6%25.8%17.2%4.0%
Adalimumab23.2%573 of 2,46820.9%24.3%22.9%24.8%3.1%
Tildrakizumab20.1%496 of 2,46814.6%25.6%19.9%20.3%2.7%
Etanercept14.7%364 of 2,46814.1%18.8%14.6%11.5%1.9%
Icotrokinra8.9%220 of 2,4687.8%6.3%9.1%12.5%1.7%
Zasocitinib6.6%163 of 2,4688.1%4.1%3.4%10.9%1.0%
Dimethyl Fumarate6.1%151 of 2,46811.7%2.4%7.5%2.9%0.9%
Certolizumab Pegol5.0%124 of 2,4683.1%6.0%7.0%4.1%0.6%
Envudeucitinib3.2%78 of 2,4685.5%2.9%0.3%3.9%0.5%
ORKA-0011.1%28 of 2,4680.3%0.0%0.0%4.2%0.1%
Orismilast0.9%23 of 2,4681.0%2.1%0.0%0.6%0.1%
ORKA-0020.2%4 of 2,4680.0%0.0%0.0%0.6%0.0%
Piclidenoson0.1%2 of 2,4680.3%0.0%0.0%0.0%0.0%
Simepdekinra0.0%1 of 2,4680.2%0.0%0.0%0.0%0.0%
Mufemilast0.0%0 of 2,4680.0%0.0%0.0%0.0%0.0%

Brands

TermAll systemsClaudeChatGPTPerplexityGeminiShare of Voice
Sotyktu1.7%41 of 2,4684.9%1.1%0.0%0.6%17.2%
Bimzelx1.6%39 of 2,4684.5%1.3%0.2%0.3%14.4%
Cosentyx1.3%32 of 2,4684.7%0.5%0.0%0.0%13.4%
Taltz1.3%31 of 2,4684.4%0.2%0.0%0.5%10.7%
Kyntheum1.1%26 of 2,4684.1%0.2%0.0%0.0%8.9%
Skyrizi0.8%19 of 2,4681.1%1.8%0.0%0.2%9.3%
Skilarence0.6%14 of 2,4680.6%0.6%0.2%0.8%6.2%
Stelara0.4%11 of 2,4680.2%1.5%0.0%0.2%6.5%
Tremfya0.4%10 of 2,4681.3%0.2%0.0%0.2%3.4%
Ilumetri*0.3%8 of 2,4680.0%1.1%0.0%0.2%4.1%
Otezla0.2%5 of 2,4680.5%0.3%0.0%0.0%1.7%
Humira0.2%4 of 2,4680.5%0.2%0.0%0.0%3.1%
Enbrel0.1%3 of 2,4680.2%0.2%0.2%0.0%1.0%
Cimzia0.0%0 of 2,4680.0%0.0%0.0%0.0%0.0%
Remicade0.0%0 of 2,4680.0%0.0%0.0%0.0%0.0%

* Includes answers using ‘Ilumya’, an alias of the tracked Ilumetri brand asset.