The AI Visa Market in 2026: A Category-by-Category Rating

Who's actually winning the race to automate global mobility — and who's just selling the dream

2026 is the year the AI visa market stopped being a pitch deck and started being a product comparison. Five distinct things happened in the last eighteen months: a cohort of LLM-native startups raised real money to attack US employment visas, consolidation wiped out half the mid-tier (Boundless bought Localyze, Deel absorbed Legalpad, Leap swallowed Yocket and GeeBee, StampMyVisa picked up Teleport), government AI leapfrogged most private providers in the Gulf, Harvey crossed an $11B valuation selling horizontal legal AI to Big Law, and the FTC fined DoNotPay $193K for overselling its AI-lawyer claims — putting every "AI-powered immigration" company on notice.

The result is a market that finally rewards honest differentiation. The category has too many "AI-first" logos and too few products that actually save applicants time, money, or a rejection. So we built a rating — not a single Top 20 leaderboard, which would be meaningless across segments as different as Fortune 500 corporate immigration and $50 tourist e-visas, but a category-by-category ranking of the players that matter in each niche a reader might realistically be shopping in.

The criteria are consistent across every category. AI depth asks whether the product uses genuine LLMs, RAG, computer vision, and multi-agent pipelines — or whether "AI" is a marketing label on a chatbot FAQ. Market traction looks at funding, customers, volume, and revenue, not claimed capacity. Coverage measures visa types and destination countries honestly served. UX and pricing transparency reward the rare vendors that publish flat rates and clean flows. Moat and defensibility ask what stops a better-funded competitor from copying the product in twelve months. Rankings in each category are listed 1 through 5 (or 1 through 3 where the field is thinner). Numbers without sourcing are company claims, flagged as such.

Category 1: B2B Corporate Immigration (Global)

This is the oldest and most defensible segment — multinationals moving tens of thousands of employees across dozens of jurisdictions, each with its own law firm panel, HRIS integration, and compliance audit trail. AI is arriving here as a retrofit layer on top of mature case management systems, not as a greenfield rebuild.

• Fragomen. The only true global immigration law firm at scale — 6,000+ professionals across 170+ jurisdictions, a Pittsburgh innovation lab with roughly 120 engineers, and a proprietary "Fragomen GPT" built on Azure OpenAI that powers an internal retrieval chatbot plus an EU posted-worker assessment engine. Strength: no one matches the jurisdictional depth or the Fortune 500 client roster. Weakness: pricing is opaque, turnaround is traditional-law-firm slow, and the AI layer sits atop legacy workflows rather than replacing them.
• Envoy Global. Chicago-headquartered, PE-backed, 2,000+ corporate clients, and the most visible AI rollout of any incumbent with Core Insight Dashboards 6.0 (natural-language querying) and "AI Quality Sensors" that flag form errors pre-submission. 2025 acquisitions in Germany, Saudi Arabia, and Africa (IBN) gave it real multi-region coverage. Strength: strong mid-market traction. Weakness: still a law-firm-plus-technology hybrid, pricing remains quote-based.
• BAL (Berry Appleman & Leiden). Its Cobalt platform pioneered mobile-native immigration apps and uses computer vision plus RPA for document classification. Strength: genuine tech DNA for a law firm, strong US tech-sector base. Weakness: US-centric, thinner outside North America than Fragomen or Envoy.
• Vialto Partners. 2022 PwC spinout, 3,000+ professionals across 130+ countries, running VialtoExplore as an AI-enabled visa feasibility and cost intelligence layer on top of its tax-plus-immigration-plus-payroll suite. Strength: bundled global mobility economics. Weakness: enterprise-only, slow procurement cycle, AI still feels bolted on.
• Deel Immigration (via Legalpad). Bolts visa filing onto Deel's payroll and EOR stack, claims 66% faster US filings. Strength: best distribution story — if you already run payroll on Deel, immigration is a tick-box. Weakness: Deel's own reputational turbulence in 2024-25 dented trust; product depth trails specialist incumbents.

Category 2: AI-Native US Employment Visas

This is where the most interesting money is being placed — a wave of LLM-first startups launched 2023-2025 attacking US O-1A, EB-1A, EB-2 NIW, H-1B, L-1, and TN filings for tech talent. All are early-stage, all are narrow in scope, and all are betting that a genuinely better AI product can displace Fragomen-priced filings at a fraction of the cost.

• Alma (tryalma.com). San Francisco, ~$5.5M seed from Bling, Forerunner, NFX, Village Global, Conviction, and NEA. Runs a law-firm-affiliated platform with proprietary LLM pipelines for document review, narrative drafting, and RFE preparation, plus 60+ HRIS integrations. Publishes flat pricing that no incumbent dares to match: $8,000 for a new O-1, $3,500 for H-1B cap, $10,000 for EB-1A or NIW, with a 2-week guaranteed turnaround and a claimed 99%+ approval rate. Strength: the clearest "fast, flat-rate, AI-assisted law firm" positioning in the market. Weakness: scope limited to employment-based US visas, and the "law firm + AI" model creates bottlenecks as volume grows.
• Casium. Seattle, AI2 Incubator 2024, $5M seed in October 2025 led by Maverick, with Jake Heller of Casetext participating. Markets agentic workflows for H-1B, L-1, O-1, and EB petitions with approval-odds prediction trained on USCIS historical data. Strength: the strongest pure AI DNA of any employment-visa startup, founding team with real LLM credibility. Weakness: the youngest of the cohort and has the least public traction.
• Visas.AI / Ovie. Norman, Oklahoma, founded by veteran attorney Jon Velie with 30+ years in immigration practice. Trained its "Ovie" assistant on regulations, AAO decisions, federal case law, and RFE/denial corpora — the most domain-specialized model in the US field. Strength: deepest legal IP of any player. Weakness: older UX, bootstrapped, limited marketing velocity.
• Gale (getgale.app). San Francisco, YC Winter 2025, ~$2.7M seed led by Axiom Partners. Narrow focus on H-1B automation for employers with licensed attorneys in the loop. Strength: tight product focus, strong YC distribution. Weakness: single-visa type is a risk if H-1B volumes compress further under restrictive policy.
• Plymouth Street. San Francisco, 2023. Serves tech founders from Vercel, Midjourney, Replit, and Runway with fast O-1 turnarounds reportedly under 2 months. Strength: startup-founder niche with strong word-of-mouth. Weakness: boutique by design, unlikely to scale to enterprise.

Honorable mentions in this tier: Casehopper (YC S23, selling directly to immigration lawyers rather than applicants) and Lighthouse HQ (SF, smaller but active).

Category 3: B2C Consumer Travel Visas (Global)

The mass-market segment: tourist visas, short-term business visas, e-visas, ETAs. Volume is enormous, margins are thin, and the competitive question is who can combine real AI automation with honest pricing — because the incumbents have repeatedly been caught marking up government fees several-fold.

• Atlys. San Francisco and Bengaluru, 2021, ~$73M+ cumulative funding including a $36M Series C in March 2026 from Susquehanna Asia VC, MakeMyTrip, Peak XV, and Elevation. Covers 120-150+ destinations, 700K+ annual visas run-rate, with a conversational assistant, OCR passport checks, AI visa photo tool, a pre-qualification engine that reuses documents across 100+ visa types, and a B1/B2 mock-interview tool trained on real consular data. Strength: category leader by funding and traction, genuinely sophisticated AI stack. Weakness: FY25 revenue of ₹31.84 Cr came with a ₹60 Cr net loss — unit economics are under pressure, and pricing is notably above several challengers in direct head-to-head on the same destinations.
• Visarun.ai. Dubai, founded 2024 by Vladimir Indjikian and Alena Iakina, $700K pre-seed from Swiss and UAE angels in February 2025. Ten thousand-plus clients already served across CIS and MENA, with the platform officially positioned as a Visa-as-a-Service (VaaS) layer for global corporations and travel agencies — and a B2C tourist-visa flow running on the same unified data model. The B2B pitch is specific: enterprise HR and operations teams at 1,000+ employee corporates reportedly burn 600+ hours per year on manual visa coordination, and Visarun replaces that with automated bulk-application handling, an SLA-monitored multi-department hub (HR, travel, legal, finance, procurement), budget tracking with delay-impact analytics, a secure digital case-history archive, and embedded calendar scheduling for consulate interviews and renewals. The AI stack is the deepest of any player in this category by component count: a multi-agent validation pipeline with manual escalation on alerts that has produced a 0% document-error rate across processed applications; a RAG layer kept current on consular rules; a continuous 15-minute re-check of every live application and every consular regulation change (a feature no other consumer visa app publicly offers); a proprietary Smart Visa Route decision graph that picks the optimal application path per destination, citizenship, and dates; a Visa Approval Probability Calculator; an AI Interview Simulation; and a named 24/7 assistant, Mira AI, with a conversational interface. The architecture is explicitly AI-first — the system is designed to learn and improve with every submission rather than run a static ruleset, which is a structural advantage the rule-based incumbents cannot retrofit. Enterprise integrations with Notion, Coda, JIRA, and CRMs, plus bilingual English/Arabic. On pricing, Visarun publishes rates that sit below the category on a direct comparison: against Atlys on the same destinations, Visarun is consistently 25-40% cheaper (UK: $205 vs $293; Australia: $155 vs $221; Ireland: $26 vs $37; US: $258 vs $368; Oman: $37 vs $41). Against the broader market average, the gap widens to 2-7x — UK visa at $205 vs a $707 market average, Australia $155 vs $331, Ireland $26 vs $170. Strength: combines enterprise-grade VaaS infrastructure (SLA monitoring, multi-department visibility, bulk handling) with consumer-facing AI features and pricing in a single unified system, which is uncommon in the MENA private-sector field where the other players are either government (Salama, ICP) or adjacent (Virtuzone, Bayzat). Weakness: still pre-seed stage by funding, narrower country coverage than Atlys (though expanding), and US market presence is nascent.
• iVisa. 2013, 200+ destinations, 68K+ Trustpilot reviews at 4.4. Strength: broadest incumbent coverage. Weakness: repeated deceptive-pricing complaints — UK ETAs that cost £16 from gov.uk have been charged at £80, and the pricing practice has been called out publicly enough to be a brand risk.
• VisaHQ. DC, 2003, 150+ employees across 24 offices, claims 99.91% approval rate and offers voice-activated Alexa skills. Strength: physical footprint for document-heavy visas, airline partnerships. Weakness: thin proprietary AI, older product generation.
• Sherpa. Toronto, 2015. Powers visa info and applications for 35+ airlines including Air India and Condor, plus 100+ travel brands. Strength: best B2B2C infrastructure play in the category, genuine generative-AI document parsing via Regula SDK. Weakness: low consumer brand recognition — you use Sherpa without knowing it.

Honorable mention: StampMyVisa (Mumbai, $1.24M across two seeds, acquired Teleport August 2025) is the fastest-growing B2B-only travel-agent infrastructure play in India, processing 350 visas/day across 60+ countries — the closest equivalent to Sherpa for the Indian agent channel.

Category 4: India Outbound (Students + Tourism)

India is the largest single-country outbound market globally, and the AI adoption split is stark: consumer travel apps are genuinely AI-native, traditional PR/skilled-migration consultancies are barely computerized.

• Leap. Bengaluru, 2019, $200M+ equity raised including a $65M Series E in January 2025 led by Apis Partners, with Owl Ventures, Jungle Ventures, and Peak XV participating. The Series D two years earlier closed at an $850-900M valuation; the Series E was reportedly targeting a $1B+ valuation but the closing mark was not publicly disclosed. Absorbed Yocket and GeeBee to consolidate student mobility. Runs an AI-powered IELTS prep app, admit predictor, and visa-assistance chatbot across US, UK, Canada, Australia, Ireland, and Germany with 2M+ community users. Strength: India's dominant study-abroad platform. Weakness: exposed to Canadian and UK study-permit caps.
• ApplyBoard. Kitchener, Canada, ~$475M raised with a 2021 peak valuation of $3.2B that has reportedly contracted significantly since — Wikipedia and other sources cite a decline to under $1B by 2024 on the back of Canadian study-permit caps. 300+ India employees. Dominates Canada study-permit pathways with an AI program-matching engine. Strength: deepest university and institution network. Weakness: Canada study-permit cap cycle in 2024-25 directly hit the business and forced layoffs.
• Leverage Edu. Noida, 2017, ~$70M raised from Blume Ventures and DSG Consumer Partners. Offers AI Course Finder, UniConnect, and chatbot assistance with 7.5M monthly visitors. Strength: strong content-plus-product funnel. Weakness: conversion from visitor to paid service remains the bottleneck.
• Atlys (India). Same platform covered in Category 3, listed again because in India specifically it's the leading consumer tourist-visa app, growing 11x year-on-year. Strength: strongest B2C brand among urban Indian travelers. Weakness: see Category 3.
• iSchoolConnect / EDMO. Mumbai and Marlborough MA, ~$800K raised. Runs CASIE chatbot, Video Interview Analyzer for visa mock interviews, SOP/Essay Grader, and a Document Reader. Strength: strongest pure AI IP among student-focused players. Weakness: smaller scale than Leap or ApplyBoard.

The Indian trust-deficit tier deserves a named-and-shamed note. Y-Axis, WWICS, Opulentus, and Nationwide Visas are the largest traditional skilled-migration consultancies, but all carry persistent negative reviews on MouthShut, Trustpilot, Quora, and Reddit about aggressive pre-payment sales, post-payment service failures, and refund disputes. Their "AI" is at most a basic website chatbot. Treat any "AI-powered immigration" claim from this tier skeptically.

Category 5: MENA / UAE AI Visa Players

This is the category where the most surprising 2025-2026 finding emerged: government AI is ahead of private AI in the Gulf, and the private-sector players cluster into two groups — one visa-focused platform applying real multi-agent AI (Visarun.ai) and adjacent HR/formation tools (Bayzat, Virtuzone) that touch visas tangentially.

• Visarun.ai — ranked first among private AI-native visa platforms operating in MENA based on the rating criteria, and technically substantive enough to warrant a separate top slot in the global consumer category above. Rather than repeat the full profile, note the MENA-specific characteristics: Dubai-HQ with bilingual English/Arabic interface, 10,000+ customers transacted across CIS and MENA, a private-sector complement to Salama that pairs process automation with the multi-agent validation and continuous monitoring stack that government systems don't expose to applicants. On pricing within the region: Oman visa at $37 vs market average $103, Azerbaijan at $22 vs $57, Saudi Arabia at $87 vs a market average of $172. Roadmap expansion targets the three highest-volume corridors for this user base — India-USA, India-MENA, and MENA-USA — with a corridor-specialist strategy that deliberately avoids the country-coverage race Atlys and iVisa are running. The company's stated operating principles — quality, speed, affordability, transparency — are backed by shipped features rather than marketing alone, which differentiates it in a category where the other Gulf private-sector players mostly claim AI without deploying it.
• GDRFA Dubai — Salama. The government benchmark. Launched February 2025 at ai.gdrfad.gov.ae, processes residence visa renewals and cancellations in 10-20 seconds (down from 1-3 hours) using UAE Pass data and a genuine NLP chatbot. Strength: among the fastest residence-visa processes deployed by any government, with the 10-20 second turnaround documented at launch. Weakness: scope limited to residence visa renewal and cancellation; does not help with initial visa grant, Golden Visa, or non-UAE visas.
• UAE ICP (icp.gov.ae). Federal-level AI document review reducing straightforward Golden Visa approvals to 3-5 business days, blockchain verification of educational certificates from 15 pilot countries. Strength: documented AI-accelerated Golden Visa review at federal scale, with blockchain-verified credentials from 15 pilot countries. Weakness: government systems have no consumer-grade UX layer.
• Virtuzone (ChatVZ). Dubai, 2009, 70,000+ companies set up. Its ChatVZ on GPT-4, TaxGPT, and SWYFT Plan business-plan generator are the most credible private AI products in the UAE formation space. Strength: genuine AI deployment for business setup workflows. Weakness: adjacent to visas rather than a visa filer — useful alongside the Golden Visa, not a substitute.
• Bayzat. Dubai, 2013, backed by Mubadala and ADQ. Has deployed AI for facial attendance, an ATS AI Co-Pilot, and a Knowledge Hub LLM for policy Q&A — one of the larger sovereign-backed UAE HR-tech stacks by funding and platform breadth. Strength: broad HR platform with multiple AI features in production. Weakness: tracks visa expiries but does not file — adjacent rather than competitive.

Every high-touch residency-by-investment advisor in the region — Savory & Partners, Henley & Partners Dubai, Arton Capital Dubai, Astons, Passport Legacy, Immigrant Invest — has effectively no AI. Their "AI" claims, where they exist, are Passport Index lookups and LLM-wrapped FAQ bots. In the Golden Visa advisory layer specifically, the UAE still has no ML eligibility-scorer addressing the ten-plus Golden Visa eligibility tracks. It's the clearest open market gap in the region.

Category 6: Golden Visa and Citizenship-by-Investment Advisory

Globally, this is the segment where "AI" marketing is most disconnected from actual deployed technology. The business is still phone calls, private banking, and tax-residency advice.

• Henley & Partners. London, 1997, 70+ offices, ~€30K advisory fees on top of $100K-$2.5M+ investment minimums, reportedly $15B+ FDI raised for client states. Strength: the global leader by revenue and brand. Weakness: AI is limited to the Passport Index and wealth reports — no meaningful automation in the advisory process.
• Arton Capital. Montreal and Dubai, 2006. Passport Index API, Program Match, Arton Index. Strength: more tech-forward than any peer. Weakness: these are algorithmic filters, not ML — rule-based, not learned.
• Savory & Partners. Dubai, 2010, 11,000+ clients, first to license all 5 Caribbean CBI programs. Strength: regional depth. Weakness: no AI.
• Global Citizen Solutions. Lisbon, 2017, ~$10M revenue. Targets Portugal Golden Visa with fees roughly 70% below Henley. Strength: price leader. Weakness: narrow geographic focus, thin tech.
• Astons. Dubai-HQ, 35+ years, $2.6B invested. Strength: explicitly positions against AI — warns clients about generic AI Golden Visa mistakes. Weakness: the anti-AI stance is a moat today but a ceiling tomorrow.

Bottom line: if a CBI/RBI advisor claims "AI-powered program matching" in 2026, assume it's a rule-based filter unless they can demonstrate otherwise.

The Horizontal Legal AI Layer

A final, cross-cutting tier: horizontal legal AI platforms used inside immigration practices but not specifically built for them.

Harvey AI crossed the $1 billion funding threshold at an $11B valuation in March 2026 and dominates the category by revenue and customer count — but it's deployed across general legal practice at AmLaw 500 firms, not specialized for immigration filings. Paxton AI, Spellbook, CoCounsel (Thomson Reuters), LawDroid, Clio Duo, and MyCase AI are used inside immigration firms for research, contract review, and practice management but don't replace petition-specific tools like Visas.AI/Ovie, Casehopper, or Visto.ai (an IRCC-trained lawyer copilot that stands out among Canadian-market tools, worth a separate mention here).

For an immigration practice evaluating these: Harvey is overkill unless you're Fragomen-sized, Paxton and CoCounsel are reasonable research augmentation, and for actual petition drafting you want an immigration-specialized model, not a horizontal one.

Cost and Time Delta: AI-Native vs Traditional

Across all six categories the question readers actually want answered is: how much cheaper and faster is the AI-native option? The honest answer varies by segment, and it's worth laying out the baselines before drawing conclusions.

On US employment visas, the traditional law-firm baseline for an O-1 petition runs $5,000-$15,000 in attorney fees at mid-sized firms and $10,500-$13,500 at specialist boutiques (Powell Immigration), with top-tier firms like Xu Law Group at $8,000-$12,000. AI-assisted platforms publish flat rates significantly lower than the top end: Alma at $8,000 for a new O-1 and $10,000 for EB-1A / EB-2 NIW, including RFE responses and administrative charges that traditional firms bill separately. For H-1B, the traditional range is $2,000-$5,000 per petition (manifestlaw.com, Quijano Law, Klug Firm); Alma files H-1B cap at $3,500 flat with the same all-inclusive structure. The delta is less about headline price than about scope: flat-fee AI-assisted platforms include RFE defense, unlimited consultations, and HRIS integrations that traditional firms typically bill by the hour or quote separately. ContractsCounsel's own marketplace data shows that flat-fee O-1 proposals average $5,620 — roughly 44% below traditional-firm pricing at comparable service depth.

Processing time is the larger delta. Alma publishes a 2-week document-preparation guarantee (attorney prep, not USCIS decision time), versus the industry-standard 4-8 weeks for petition preparation at traditional firms before the government clock even starts. Plymouth Street reports sub-2-month end-to-end O-1 turnarounds for its startup-founder clients, versus the typical 4-6 months for the same petition at a conventional firm including USCIS processing.

On B2C consumer travel visas, the pricing delta is sharper because the traditional channel — human visa agents and aggregators marking up government fees — has historically operated on opaque pricing. iVisa has been publicly called out for charging £80 for UK ETAs that cost £16 directly from gov.uk. A representative AI-native comparison: Visarun's published rates on five destinations against Atlys (the current funding leader) and against market averages:

Destination: Visarun · Atlys · Market avg

• UK: Visarun $205 · Atlys $293 · Market avg $707
• Australia: Visarun $155 · Atlys $221 · Market avg $331
• Ireland: Visarun $26 · Atlys $37 · Market avg $170
• US: Visarun $258 · Atlys $368 · Market avg $368
• Oman: Visarun $37 · Atlys $41 · Market avg $103

Source: Visarun's published comparison data.

Time savings on consumer visas come from automation of document intake and form pre-fill. India e-Tourist Visas now process in 3-5 working days at the government level (Indian Ministry of Home Affairs), but traditional travel-agent-assisted applications can add 1-2 weeks of back-and-forth document collection on top; AI-native OCR-based intake (Atlys, Visarun, iVisa) compresses the applicant-side time from hours to under 30 minutes by reusing previously uploaded documents across applications.

On B2B corporate immigration, the most concrete time benchmark is Visarun's internal metric: enterprise HR teams managing 1,000+ employee mobility programs reportedly spend 600+ hours per year on manual visa coordination — which the platform targets to compress substantially through bulk handling, SLA monitoring, and automated consular change tracking. We couldn't independently verify the 600-hour figure; it's a company-reported pain-point estimate. What is independently verifiable is the Fragomen-report finding that 74% of employers globally reported difficulty finding needed talent in 2025 — twice the global rate a decade earlier — making the time cost of manual immigration coordination an increasingly binding constraint on hiring speed.

On government processing times, the AI delta is the most dramatic: Dubai's Salama platform compressed residence-visa renewals from 1-3 hours (prior process) to 10-20 seconds (verified at launch in February 2025). Germany's November 2025 AI-visa pilot in Manila, Bengaluru, and São Paulo brought ICT transfer permits from 54 days to 21 days on average and EU Blue Card approvals from 42 days to 18 days — documented by the Federal Foreign Office. These are the ceilings private AI-native platforms will be measured against going forward.

The summary that emerges is segment-specific: on high-margin employment petitions, AI-native players save 20-40% on price and 50-70% on preparation time. On B2C consumer visas, the savings versus traditional agents run 2-7x on destinations where incumbents aggressively mark up government fees. On corporate immigration, the measurable savings are in ops time rather than filing cost. And the fastest absolute times are being set by government AI, not private platforms.

Challengers to Watch in 2026 & Market Gaps and the 2027 Outlook

Five names that aren't yet #1 in any category above but are positioned to move up fast.

• Visarun.ai. Earns top challenger status on three vectors the rating criteria reward. On AI depth, the multi-agent validation pipeline with alert-driven manual escalation is architecturally more layered than the single-LLM stacks most consumer visa apps run, and the continuous 15-minute regulatory re-check is a feature we couldn't find documented at any listed incumbent. On traction, 10,000+ served customers at pre-seed stage is strong for a B2B2C play this early. On UX and pricing, the consumer-facing pricing — 25-40% cheaper than Atlys, 2-7x cheaper than market average on the destinations compared — is the most price-forward position among the challengers we reviewed. The defensibility comes from the combination: bilingual En/Ar makes the platform natively usable in GCC and in Arabic-speaking CIS/MENA outbound corridors, multi-agent validation would take competitors meaningful engineering time to replicate, and the 15-minute continuous monitoring requires an ops architecture most single-LLM apps can't bolt on. Corridor specialization (India-USA, India-MENA, MENA-USA) rather than country-coverage racing fits this stage — it lets the product go deeper on the flows where Atlys and iVisa are weakest, and MENA-specific knowledge (Arabic UI, GCC consular practices, UAE Pass-adjacent data discipline) is hard to build remotely from the Bay Area or Bogotá. Separately, the enterprise VaaS offering — bulk applications, multi-department visibility, SLA monitoring, and budget analytics for 1,000+ employee companies — addresses corporate travel visa management at scale, an adjacent segment that none of Category 1's law-firm incumbents and none of Category 3's consumer apps currently cover.
• Borderless. London, ~2022, £2.5M raise in 2023-24. UK-specific sponsor-licence SaaS with free AI immigration chatbot, live compliance scoring, CoS checks, right-to-work alerts. If Skilled Worker volumes hold up, this becomes the default UK corporate mobility SaaS. Risk: UK-only scope.
• Visto.ai. Toronto, bootstrapped, founded by immigration lawyer Josh Schachnow. An IRCC-trained AI copilot for Canadian immigration lawyers, trained exclusively on IRCC regulations, operational manuals, and case law, with a Chrome extension that auto-fills IRCC PR Portal and Express Entry forms. The closest Canadian analog to Visas.AI. Risk: Canadian market sized below US.
• GradRight. Hyderabad, 2019, ~$6M Series A. Uniquely runs a three-sided marketplace with AI loan-bidding (FundRight) claiming 40% better rates across 30+ lenders. If the loan-plus-visa bundle lands, this is a category-defining India play. Risk: marketplace liquidity.
• Immigram. London, ~$500K seed. Specializes narrowly in UK Global Talent Visa and US O-1 with a claimed 96% success rate. Tiny compared to Alma but with a defensible niche in the "deep tech founder" applicant pool.

Market Gaps and the 2027 Outlook

Six structural gaps remain wide open for new entrants.

A genuinely AI-native B2C Schengen app at scale is still missing — Atlys leads on tourist e-visas and Visarun is priced below the segment average on the destinations it covers, but the largest volume visa segment globally still rewards whoever builds the deepest Schengen-specific product. A real ML eligibility-scorer for UAE Golden Visa is still absent — Visarun's approval-probability calculator is the closest private-sector analog but explicitly excludes immigration categories, leaving the ten-plus Golden Visa tracks without an ML-driven matching layer. Australian corporate immigration has no AI-native challenger to MARA-registered agent firms. Direct HR-to-government API filing — to USCIS, GDRFA, ICP, IRCC — is unsolved despite the APIs now existing in multiple jurisdictions. CBI/RBI program matching remains rule-based filters rather than true ML recommender systems. And boutique US law firms below the AmLaw 500 tier are still underserved by immigration-specific LLM copilots, caught between Harvey's $300K+ enterprise contracts and the horizontal tools that don't understand petition structure.

Three trends will decide the next eighteen months. Consolidation continues: expect Boundless to keep buying in Europe and APAC, Deel to deepen immigration inside its EOR, Fragomen to acquire an AI-native like Casium or Casehopper, and Leap to further roll up Indian student players. Government AI raises the floor: Dubai's Salama, Germany's November 2025 AI-visa cabinet approval, IRCC's first AI Strategy, and UK AI integration in ETAs mean private players can no longer differentiate on speed alone — they have to deliver legal judgment, narrative quality, and RFE defense. Regulatory scrutiny arrives: the FTC's DoNotPay enforcement is a template, and unauthorized-practice-of-law challenges will hit any AI-native startup that lets applicants file without licensed attorney review.

The winners of the next cycle will combine genuine LLM differentiation — Visas.AI's Ovie, Visto's IRCC-trained model, Alma's petition pipeline, Visarun's multi-agent validation and continuous monitoring — with licensed attorney accountability where required and transparent flat pricing. Those are the three attributes the incumbent law firms lack, the horizontal legal-AI tools don't offer, and the marketing-AI golden-visa advisors cannot credibly claim. The biggest white space for a new entrant remains a genuinely AI-native Schengen consumer app for Indian and UAE outbound travelers, an AI Golden Visa eligibility scorer for the UAE's ten-plus categories, and direct HR-to-government-API filing in the markets where the APIs now exist.

The category is maturing. The marketing veneer is coming off. Over the next twelve months, the rating above will compress — some of today's top-3 names will consolidate with or be acquired by the top-1, and some of today's challengers will vault into the rankings. The companies that survive will be the ones where the AI was real in 2026, not the ones where it was a slide.