Summary

  • Fiberty is not an assumed alias. A provincial incorporation notice, a national communications licence, the company’s customer contract, a trademark bulletin and internet-number records all connect the brand and operating domain to ARTIFICIAL INTELLIGENCE RANDOM ACCESS LABORATORY SAS.
  • Permission and deployment are different things. The licence authorises a broad range of wired and wireless services, while a Río Negro infrastructure register supplies unusually concrete evidence of fibre service at Bariloche’s productive and technology park. Neither document proves capacity, coverage or uptime at any particular address.
  • The company holds AS273846 and an IPv6 allocation, but public route observation at the research cut-off showed neither the autonomous system nor its allocated IPv6 block being announced. That does not disprove an operating access network; it does mean buyers cannot treat the ASN as evidence that Fiberty presently controls the public route from subscriber to internet.
  • Fiberty’s strongest retail differentiators—buried fibre, local support and simple 100, 300 and 500-megabit offers—are accompanied by material unanswered questions about upload speed, aggregation, restoration targets, routing, power resilience and the boundary between line performance and in-home Wi-Fi.
  • For an SME, the right buying test is not a speed-test screenshot. It is an address-specific design, a wired peak-hour trial, a named escalation path, a measurable SLA, a map of genuinely diverse failure paths, and evidence that each promised layer is under accountable operational control.

A corporate name that sends the reader in the wrong direction

Begin with a small trap. ARTIFICIAL INTELLIGENCE RANDOM ACCESS LABORATORY SAS sounds like a research venture built around machine learning or experimental computing. The public legal record says something more prosaic and more useful. In the Río Negro official bulletin of August 27, 2020, the company was constituted in San Carlos de Bariloche with Ricardo Marcelo Pincin as its shareholder and legal representative, Moreno 24 as its seat, and a stated purpose centred on telecommunications and information-and-communications-technology services. The purpose covered fixed and mobile, wired and wireless, national and international services, with or without the company’s own infrastructure. The name, in other words, is not evidence of a laboratory. The original corporate purpose is evidence of a telecoms vehicle.

The next link in the identity chain arrived in December 2021. ENACOM Resolution 2101/2021 granted the exact legal company a licence to provide ICT services and entered it for value-added internet access, data transmission and subscription broadcasting over physical or radio links. Fiberty’s published customer terms then close the gap from regulator to storefront: they identify the site as belonging to ARTIFICIAL INTELLIGENCE RANDOM ACCESS LABORATORY S.A.S., give CUIT 30-71689397-5 and the Moreno 24 address, abbreviate the provider as AIRA LAB SAS, and say that internet service is supplied under Resolution 2101/2021.

There are two more, independent joins. An INPI trademark bulletin associates the word FIBERTY with ARTIFICIAL INTELLIGENCE RANDOM ACCESS LABORATORY in class 38, the telecommunications class. And LACNIC’s current registrant record gives the exact legal name, the Bariloche address, Ricardo Pincin as legal representative and a fiberty.com.ar contact, while attaching internet-number resources to that registrant.

That is a stronger identity bridge than a logo, a social account or an uncorroborated directory listing. It permits this article to analyse Fiberty as the public operating brand of the assigned company without silently substituting a parent, an affiliate or a similarly named network. It also sets a hard boundary. A separate organisation called AIRA Comunicaciones appears later in the public network evidence. Its similar acronym is not proof of common ownership, and this analysis does not treat it as the same company.

The operating timeline is now long enough to matter. Incorporation dates to August 2020; the national licence to December 2021; the trademark record to 2023; internet-number allocations to June 2024; and the public product, ordering and enterprise pages were active at the July 2026 research cut-off. That sequence proves sustained administrative and commercial activity. It does not, by itself, disclose subscriber numbers, turnover, staff size or profitability, and none of those is estimated here.

What Resolution 2101 permits—and what it pointedly does not prove

Fiberty’s licence is broad because Argentina’s licensing framework is broad. Law 27.078 says licences may enable fixed or mobile, wired or wireless, national or international services, with or without the licensee’s own infrastructure. Resolution 2101 mirrors that language. It also registers internet access, data transmission and subscription broadcasting. This is legal permission to operate across several technical designs, not a certificate that any one of them has been built.

The resolution includes a particularly useful warning. It says the licence does not oblige the state to guarantee the availability of radio spectrum; any frequency authorisation or permit must be processed separately. That sentence is a miniature lesson in telecoms due diligence. A company can be permitted to provide wireless service without holding the particular spectrum required for a contemplated link. It can be licensed to operate with its own infrastructure without having completed civil works. It can be licensed to operate without its own infrastructure while buying wholesale capacity from someone else.

It can register data transmission without proving a redundant route at a customer’s building.

Permission is therefore the first rung of evidence, not the last. For Fiberty, the useful ladder looks like this:

  1. Legal capacity: the provincial notice proves the company exists and was formed for telecoms work.
  2. Regulatory permission: Resolution 2101 proves it may provide the registered services, subject to the rest of the regulatory regime.
  3. Commercial commitment: Fiberty’s terms identify the licensed company as the contracting provider.
  4. Physical presence: an independent infrastructure register and the company’s address-specific sales process show that access is more than a generic national licence.
  5. Active capacity: only plant records, commissioned circuits, utilisation data and tests can show what capacity exists at a given place and hour.
  6. Operational accountability: tickets, maintenance notices, escalation records, restoration performance and credits show whether the service is actually managed as promised.

The last two rungs matter most to a subscriber and are least visible in public. That is not unique to Fiberty. Private operators rarely publish detailed feeder maps or utilisation charts. But the absence changes what a buyer may responsibly conclude. The licence is decisive evidence that Fiberty is authorised; it is not a substitute for a site survey, an acceptance test or an SLA.

PITBA is the strongest public clue that fibre reached a real operating surface

The most valuable deployment evidence does not come from Fiberty’s advertising. It comes from Río Negro’s register of industrial and logistics parks. The provincial entry for the Parque Productivo Tecnológico e Industrial Bariloche, or PITBA, describes a 319-hectare site on Route 40, with 222 functional units and the first of four infrastructure stages completed. Under telecommunications, it says fibre is available and names FIBERTY as the service provider.

This is unusually specific. It ties the brand to a defined facility about seven kilometres from central Bariloche and five kilometres from the airport. It demonstrates an operating surface where connectivity is part of a larger bundle of industrial infrastructure—electricity, water, gas, roads, access control and security—rather than a promotional pin on a city map. For an ISP whose enterprise page targets industry, logistics, health, technology and multi-site customers, the park is a credible reference point.

Yet even this record has limits. It does not say how many PITBA units are connected, when each circuit entered service, how much feeder capacity reaches the park, whether there are two physically independent routes, where active equipment is powered, or whether Fiberty owns or leases every segment. It names a provider; it does not publish a network acceptance dossier. Nor should the park listing be extrapolated into blanket city coverage.

Fiberty’s home page claims buried fibre in Bariloche’s neighbourhoods and centre. Its ordering page is more revealing because it makes availability conditional. A prospective customer selects a neighbourhood or building, supplies an exact address or map position, and, where service has not arrived, may register to be notified as coverage expands. The sales workflow therefore concedes what every fixed-access operator knows: a city-level claim is not an address-level commitment.

Taken together, PITBA and the order form are a sensible pair of proofs. The independent register shows that at least one substantial site recognises Fiberty as a fibre provider. The address gate shows the operator is not pretending that every location is immediately serviceable. What remains unavailable is a public polygon map or premise-level coverage dataset. A purchaser should treat “Bariloche” as the market, not as a guarantee, until the provider issues a written feasibility result for the exact entrance, floor, unit and equipment room.

The customer journey is also a map of the network’s economics

Fiberty’s retail funnel begins before a packet moves. The order form asks whether the address is a neighbourhood or a building; presents plan choices; collects contact, identity and installation data; offers automatic debit; and says a dual-band Wi-Fi router is included. It also offers to take a customer’s satellite antenna in exchange for a waived installation charge. Each step reveals something about the operating structure.

The published retail plans are deliberately simple: 100, 300 and 500 megabits. The 100-megabit plan is labelled exclusive to buildings. That condition is not an engineering specification, but it is economically suggestive. Entering a multi-dwelling building can spread the cost of obtaining permission, bringing in a feeder, installing vertical distribution and dispatching technicians across several potential subscribers. Restricting the entry product to that setting may help Fiberty recover civil and installation costs without offering the same economics to an isolated house. This is an inference from the offer structure, not a company explanation.

The precise address and map pin matter because the costly part of a local fibre network is not the national licence. It is the last physical path: available duct, road crossing, building permission, optical budget, drop length, safe equipment placement and technician time. Fiberty’s contract assigns customers responsibility for obtaining required permissions and making spaces available for equipment and cable runs. It gives the customer five days after installation to test the service and raise written objections. It also treats some supplied equipment as a loan that must be returned on cancellation.

Those clauses convert the abstract word “coverage” into a sequence of practical gates.

There is a second boundary inside the premises. Fiberty markets “fibre” but supplies a Wi-Fi router, and most subscribers experience the service through that radio link. The difference matters. Ofcom’s consumer engineering guidance explains that walls, metal, water, interference, router placement, client capability and the number of connected devices can all affect Wi-Fi performance; it recommends wired Ethernet when a reliable measurement is required. That guidance is British, not Argentine regulation, but the radio and Ethernet distinction is universal.

For Fiberty, a proper installation acceptance test should therefore have two records. The first is a wired test at the provider’s hand-off or supplied router, showing download, upload, latency, jitter and loss at several times. The second is a room-by-room Wi-Fi check, clearly treated as an in-premises design problem. A 500-megabit access line cannot guarantee 500 megabits through every wall to every old handset. Conversely, poor Wi-Fi should not be allowed to conceal congestion in the access or upstream network. The customer needs a repeatable way to separate the two.

Fiberty promotes optional mesh coverage, which is a reasonable response to the in-home problem. But “dual band” and “mesh available” do not disclose the router make and specification, Wi-Fi generation, Ethernet-port speed, management policy, firmware-support term, bridge-mode availability or whether the customer may use its own equipment. Those details become important at 300 and 500 megabits and indispensable for an SME that wants its own firewall.

Three speed labels conceal several possible architectures

“Fibre” describes the physical medium, not the full capacity architecture. Fiberty does not state on the reviewed retail pages whether its access system is GPON, XGS-PON, active Ethernet or a mixture. It does not publish split ratios, optical budgets, feeder utilisation or the size and redundancy of the uplinks behind each service area. That is not proof of a weak design. It means the public evidence cannot translate a headline rate into a contention pattern.

The distinction is material. ITU-T’s generic optical-media architecture describes a passive optical distribution network as a point-to-multipoint tree that can include splitters, combiners and filters. In such a design, several customer endpoints can share an optical feeder and an optical-line terminal port. ITU-T’s current XGS-PON recommendation discusses physical and logical split ratios precisely because sharing choices affect reach, loss budget and economics. Those standards explain possible architectures; they do not establish which one Fiberty uses.

That uncertainty changes the questions. A household choosing between 100 and 500 megabits may care mainly about price and evening performance. A hotel, clinic, software office or manufacturer should ask:

  • Is the quoted bandwidth best-effort, committed, or guaranteed?
  • Are upload and download rates symmetric?
  • What aggregation or oversubscription ratio applies at the PON port, access switch and upstream edge?
  • Does the “dedicated” enterprise product mean a dedicated fibre pair, a dedicated VLAN, a committed information rate over shared transport, or merely a non-residential contract?
  • Where is the service demarcation, and which equipment is covered by the SLA?
  • What peak-hour latency and packet-loss targets apply to named test destinations?
  • Does a 500-megabit port have gigabit Ethernet all the way through the supplied device?
  • Is IPv6 native, delegated to the customer and supported on the same terms as IPv4?

These are not exotic requests. Argentina’s updated Regulation 733-E/2017 says internet providers must maintain information about speed, link quality, service nature and guarantees, and must make available, on request, plan characteristics, download limits, guarantees, aggregation or overselling rates, quality indicators, restoration times, link availability and traffic-management measures. It also requires a service label containing price and performance information.

Fiberty’s terms refer to characteristics in the service request and accompanying label, which is consistent with that framework. The public pages reviewed, however, do not expose the complete set of quantitative fields. The sensible conclusion is not that the information does not exist. It is that a buyer should obtain the actual label and proposal before comparing Fiberty with another operator, and should preserve both documents as the acceptance baseline.

The quiet ASN: a useful resource, but not evidence of today’s subscriber path

Internet-number records add an intriguing layer. LACNIC assigned AS273846 to ARTIFICIAL INTELLIGENCE RANDOM ACCESS LABORATORY SAS on June 6, 2024. The associated registrant record also attaches the IPv6 block 2803:4550::/32 to the company. The legal name, address, representative and fiberty.com.ar contact make the identity match direct.

An autonomous-system number can give an operator independent routing policy, the ability to connect to more than one upstream, clearer control of portable network resources and a path towards local peering. An IPv6 /32 is ample address space for a regional access deployment. But allocation is capacity to act, not proof that the act has occurred.

At the frozen observation time, RIPEstat’s routing-status view for AS273846 reported zero IPv4 and zero IPv6 prefixes, no observed neighbours and no route visibility among its route collectors. A separate prefix view for 2803:4550::/32 marked that allocated block as not announced. The route-origin check returned no validating authorisations for that prefix. In plain terms, the public observers used here did not see Fiberty’s own ASN carrying its own IPv6 allocation on the global routing table.

That finding should not be stretched. It does not show that Fiberty has no customers, no fibre, no upstream connectivity or no working internet service. A retail ISP can number customers from a wholesale provider’s space, use private addressing behind carrier-grade translation, or operate access and transport under another network’s public route. Route collectors also have limited vantage points. The correct conclusion is narrower: AS273846 cannot presently be used as public proof of Fiberty’s active capacity, multihoming or subscriber egress.

There is also a security watchpoint rather than an accusation. LACNIC’s explanation of RPKI describes a Route Origin Authorisation as the cryptographic statement that a particular autonomous system may originate a prefix. Because Fiberty’s /32 was not being announced, the absence of a visible authorisation did not create an invalid live route in this observation. Before activating that block, however, the operator would be expected to align its announcement and authorisation so networks performing route-origin validation do not reject it.

For a technically mature buyer, the ASN is therefore a watchable option value. If AS273846 begins announcing the /32 or future IPv4 resources, if valid route authorisations appear, and if two genuinely independent neighbours become visible, the public evidence of routing control will strengthen. Until then, a procurement team should ask Fiberty to identify the actual customer egress, address owner and upstream arrangement for the circuit being quoted rather than writing AS273846 into a risk memo as if it were already the live network.

A public portal exposes one dependency chain, but not the access network

Fiberty’s navigation links its “Oficina Virtual” to a numeric IPv4 address and port rather than to the fiberty.com.ar hostname. That small implementation detail permits a bounded network check. The linked address is 45.166.178.51. LACNIC’s address record places it inside 45.166.176.0/22, registered to IRACHETA ANDRES (AIRA COMUNICACIONES), not to ARTIFICIAL INTELLIGENCE RANDOM ACCESS LABORATORY SAS. RIPEstat’s network view observed the more specific 45.166.178.0/24 originated by AS267730.

The resemblance between “AIRA LAB” and “AIRA Comunicaciones” is tempting and insufficient. The public registrants and named contacts differ. No ownership claim is made. What the evidence supports is that Fiberty’s public site points customers to an application address carried in another registrant’s network.

The current route data adds one more link. RIPEstat observed three IPv4 /24s originated by AS267730, including the portal’s /24, and its neighbour view showed AS7049 on the observed path side. LACNIC registers AS7049 to Silica Networks Argentina S.A.

This is evidence about the portal path, not the subscriber path. It would be reckless to claim from one application address that Fiberty’s household or PITBA traffic uses AIRA Comunicaciones or Silica in the same way. The address may represent hosting, business software or a management service completely separate from broadband egress. Its value is methodological: it demonstrates why a buyer must distinguish the access provider named on the contract, the owner of an application address, the origin network visible in BGP and the wholesale carrier beyond it.

The portal link also deserves a security question. On the reviewed site it uses the plain HTTP scheme with a numeric address and non-standard port. That does not prove credentials travel without encryption; the destination may redirect or establish protection in another way. It does mean a customer should verify the live redirect chain, certificate, encryption, session controls and support ownership before using the portal, rather than assuming those controls from the Fiberty brand.

Bariloche has a local exchange; public evidence does not yet put Fiberty on it

Wholesale access is often where a small ISP’s economics and resilience are decided. Buying transit is faster than building a national backbone, but price, capacity and fault concentration depend on the available carriers. Local peering can keep some traffic close, reduce unnecessary transport and create another operational community. Bariloche has such an asset.

CABASE’s IXP Bariloche page says the exchange opened in 2014, has IPv4 and IPv6 peering subnets and lists seven connected participants: ALTEC, BariNet, Coopetel, NetPatagon, Quasar, Silica Networks and Baritel. Fiberty is not in that published participant list. The absence does not prove Fiberty has no private interconnection, no reseller relationship or no route to cached content. Nor does IXP membership alone guarantee a resilient internet service. It does identify a local interconnection option against which Fiberty’s routing strategy can be tested.

PeeringDB’s API query for AS273846 returned no network entry at the evidence cut-off. Again, that is an absence in a voluntary public directory, not proof of no peering. Combined with the dormant ASN, however, it leaves the public record unable to answer four questions: where Fiberty exchanges local traffic, which networks supply full transit, whether those paths are physically diverse, and how traffic fails over.

Argentina also has a federal wholesale network. ARSAT’s open-data page for REFEFO services describes a monthly dataset of customers in service by province, but it does not establish Fiberty as one of those customers. The existence of that wholesale layer is relevant to the market; it is not a licence to invent a Fiberty–ARSAT contract. Any claim about a particular supplier needs a contract, provider statement, route record or commissioned-circuit document.

For an SME, “two upstreams” is not enough. Two commercial names can share the same duct, long-haul fibre, power feed, exchange room or carrier further north. The procurement question is whether the two paths remain disjoint at the points likely to fail. A meaningful answer should identify the building exits, street routes, aggregation sites, long-haul carriers, hand-off locations and convergence points. If that information is sensitive, it can be supplied under confidentiality and reduced to a written diversity warranty.

Buried fibre changes the failure modes; it does not abolish them

Fiberty’s most distinctive consumer message is that its fibre is underground. The home page says buried plant resists wind, snow and storms, avoids saturated poles and dangling cable, and produces a cleaner streetscape. These are company claims, not independently measured availability results. They are nevertheless coherent with a real design choice.

Underground plant can reduce exposure to falling branches, pole loading and direct weather contact. It can also be expensive to extend and slow to repair when fault location, excavation permission, traffic control or access to a chamber is required. A buried route remains vulnerable to civil works, water ingress, crushed duct, poor splicing and shared chambers. More importantly, the fibre is passive only between active endpoints. Optical-line equipment, switches, routers, portal servers and customer devices still need power and maintenance.

This is why “soterrada” should trigger a more precise resilience conversation rather than end it. A customer should ask:

  • Which portion is buried: feeder, distribution, drop, or all three?
  • Are there diverse ducts, or two fibres in the same trench?
  • Which chambers and road crossings are shared?
  • Where are the powered aggregation points, and what battery or generator autonomy do they have?
  • How is a break located, and what spare cable, closures and optics are held locally?
  • Are emergency excavation arrangements in place?
  • Does the upstream path leave Bariloche on more than one physical corridor?

The company’s public pages do not answer those questions. PITBA’s infrastructure record proves fibre availability at a serious site, but it does not publish route diversity. The retail claim establishes the intended physical form, but it does not publish measured failure rates. A good operator should be able to answer without revealing every exact coordinate.

The same discipline applies to power. Fiberty’s terms exclude liability for interruption caused by electricity or another external service. That contractual allocation does not make power irrelevant to service design. It makes backup design more important to the buyer. A household may accept that the router stops when the home loses power. A clinic, hotel or factory may need customer-side UPS capacity, backed provider equipment and a second access technology that does not share the same energy dependency.

Pricing is an acquisition system, not just a monthly number

Fiberty’s public pricing reveals a carefully layered acquisition structure. The plan page displayed list prices and a 33 per cent automatic-debit discount for 100, 300 and 500 megabits, with a further Banco Patagonia amount. It labelled those promotions valid through June 30, 2026. Because this article is dated July 19, those numbers are historical offer evidence, not asserted current prices. A purchaser should obtain a fresh written quote rather than rely on a page whose displayed validity date had passed at the research cut-off.

The architecture matters more than the peso figure. A large discount for automatic payment reduces collection friction and late-payment risk. A bank-linked discount can shift part of the acquisition economics into a commercial partnership. Restricting the lowest tier to buildings can improve the ratio between installation cost and reachable subscribers. Offering installation relief for a surrendered satellite antenna directly targets a rival access technology and lowers the cash cost of switching.

That last tactic sits in a rapidly changing market. ENACOM’s April 2026 market update reported 12.99 million fixed-internet accesses nationally, 6.2 million fibre accesses and 750,000 satellite activations, up from 452,000 in December 2025. Those are national aggregates based on provider declarations, not Fiberty figures and not a Bariloche market share. They explain why a local fibre operator would treat the satellite dish as a competitive object rather than a curiosity.

The contract adds less visible price components. It allows connection charges, monthly fees and other items specified in the service request; passes through taxes and sector-specific charges; applies interest to arrears; and can charge for unreturned loan equipment. Moving service is subject to technical availability. A customer may request immediate cancellation, but must return equipment in working condition or face a charge. The default contract term is twelve months if the service request does not specify another period, with automatic renewal language and separate provisions for early termination and non-payment.

This creates at least four prices:

  1. the advertised list rate;
  2. the promotional rate while conditions are met;
  3. the installed first-year cost, including activation, wiring, mesh or unusual works; and
  4. the exit or relocation cost, including equipment, lost discounts and a new installation elsewhere.

For households, the comparison should be made over twelve months with expired promotions removed. For SMEs, price should be normalised per protected site and per committed megabit, not per headline megabit. A cheap 500-megabit best-effort circuit and a costly 100-megabit committed circuit are different products. A second route that shares the first route’s trench is not redundancy, however attractive the bundle price.

Support is the product after the light turns red

Local support is Fiberty’s second major promise. The current retail page lists general and commercial attention on weekdays, technical support every day from 10:00 to 22:00, and a home-service response time of up to 72 hours, with the warning that some visits may be charged. The enterprise page goes further, claiming direct contact with senior technicians, monitoring, rapid response and 24/7 support for corporate operations.

The customer terms describe another layer: claims by telephone or email on weekdays, a maximum of three business days for service deficiency or interruption, five business days for billing and ten for other claims. They promise reactivation within 24 hours after payment is proved or after an improper suspension. For internet unavailability, the terms say the provider responds with a proportional credit only when interruption at the network termination point exceeds 72 hours, continuously or intermittently, in a calendar month, and only for the time beyond that threshold.

Some of these statements can coexist: a support desk may operate longer than the formal claims desk; a technician may attend within 72 hours while the regulated claim deadline is three business days; an enterprise contract may negotiate a tighter SLA. But the pages do not turn those distinctions into one obvious escalation map. A subscriber facing an outage needs to know which clock starts, what counts as acknowledgement, what channel creates a ticket, when a field visit becomes chargeable, and who can authorise a credit.

The updated customer regulation sets the three-business-day resolution limit for deficiency and interruption claims and states that service-condition failures may create a right to compensation determined by the authority. It also requires providers to publish complaint procedures and restoration information. Fiberty’s terms reproduce several of those deadlines. The harder operational question is performance: how often are they met?

The reviewed company pages do not provide a public status page, incident archive, monthly availability report or post-incident analysis. That bounded absence is not evidence that Fiberty has suffered a particular outage, nor that it conceals incidents. It means an outside reader cannot measure incident frequency or restoration performance from the public record used here.

An SME should ask for anonymised evidence covering the previous twelve months: availability by service area; ticket volume; median and high-percentile time to acknowledge and restore; planned-maintenance notice compliance; number and duration of upstream, power and fibre faults; and credits actually issued. A hotel should test the escalation path on a weekend. A clinic should know whether 22:05 falls into a staffed enterprise arrangement or an after-hours queue. A retailer should know whether a failed payment terminal is treated as a generic internet ticket or a priority business incident.

Support proximity can be a genuine regional advantage. It becomes defensible only when the named people, hours, spares, telemetry and authority to act are organised behind the promise.

The enterprise offer is attractive precisely where it is least quantified

Fiberty’s corporate page reads like the expansion path of a local access operator. It claims buried dedicated and symmetric fibre, guaranteed bandwidth, custom SLAs, public IP addresses, VPNs, segmentation, balancing, redundancy, remote-site links, firewalls, filtering, traffic priority, senior support and monitoring. It also offers structured cabling, LAN and WAN design, managed switching and routing, high-performance Wi-Fi, cloud services and technology consulting.

These are company claims, not specifications. The page publishes no standard enterprise price, committed rate, availability percentage, latency target, repair time, service-credit schedule or definition of “dedicated.” That is not necessarily a defect: bespoke networks require bespoke proposals. It does place almost all the value in the contract and design appendix rather than on the marketing page.

PITBA makes the offer more than hypothetical. A productive and technology park contains the sorts of customers that need site-to-site connectivity, cameras, stock systems, hosted applications and operational continuity. Yet those workloads expose a local ISP to harder demands than residential streaming. A plant may need public addressing, inbound access, deterministic firewall changes, logs, time synchronisation and maintenance coordination. A health customer may care about data handling and audit trails. A hotel may need guest isolation, capacity that follows occupancy and Wi-Fi coverage across construction hostile to radio.

The enterprise proposal should therefore divide responsibility into layers:

  • Civil and optical layer: route, duct, fibres, splitters if used, splicing, optical levels and repair ownership.
  • Access equipment: termination device, switch or router, software version, spare policy and power.
  • IP service: address family, address ownership, translation, routing, committed rate, queues and traffic policy.
  • Security: firewall administration, logging, segmentation, remote access, change approval and incident notification.
  • Application dependency: DNS, VPN, cloud, portal and third-party service responsibilities.
  • Support: monitoring boundary, ticket priority, escalation names, response, workaround, restoration and credits.

If Fiberty supplies both the access line and the customer’s internal network, the arrangement can reduce finger-pointing. It can also deepen lock-in and concentrate operational risk. The buyer should retain configurations, asset lists, administrator access, diagrams and an exit plan. “One throat to choke” is useful only if that throat owns the authority, documentation and spare parts required to restore the service.

Security and compliance begin at the order form

Fiberty’s order workflow asks for a name, email, telephone number, identity number, address, date of birth and a preferred way to transmit payment details. That is enough personal and financial context to make data handling part of the product. The terms say customer data will not be used for purposes beyond the contract and promise compliance with Law 25.326 and the national Do Not Call register. Law 25.326 protects personal data held in public or private records and gives data subjects rights around the information stored about them.

The reviewed order, plan, corporate and terms pages do not present a detailed privacy notice explaining retention periods, processors, portal hosting, access controls, deletion workflow or incident notification. That statement is limited to those pages; another document may exist or be supplied during contracting. A customer should request it before sending payment information through email, messaging or telephone.

The enterprise page’s security menu—firewalls, filtering, VLANs, managed routing and monitoring—addresses useful controls, but labels are not an assurance case. The proposal should identify vendor and support status; who holds administrative credentials; whether management uses multi-factor authentication; how configuration changes are approved and backed up; how logs are retained; whether customer networks are isolated by design; how denial-of-service attacks are handled; and how a compromised customer is contained without destabilising others.

Routing security is another layer. Fiberty’s allocated IPv6 block was not active in the observed global table, and no authorisation was found for the contemplated origin at the cut-off. That is a readiness issue for future activation, not evidence of a current hijack. The company should create and maintain correct route-origin authorisations when it announces its own space, publish accurate routing contacts, filter customers and upstreams appropriately, and test changes before moving production traffic.

The numerical office-portal link deserves a mundane control review too. Buyers should confirm that the portal enforces modern encrypted transport, has a certificate users can validate, protects credentials and payment data, expires sessions, records administrative actions and has a documented recovery route. A browser padlock is only one control; the point is clear ownership of the whole customer-management surface.

Finally, the contract’s limitation of liability and exclusion for external power or force majeure should be read alongside operational requirements. A legal cap does not quantify cyber resilience. An SME that cannot tolerate a day offline must engineer its own continuity—secondary access, tested failover, local backups, independent power and manual procedures—even when the primary provider offers a strong SLA.

Switching costs accumulate in cables, addresses and habits

Broadband looks easy to switch because the monthly product is described by one number. The actual cost accumulates in physical and organisational layers.

For a household, Fiberty may install cable and a loaned router. The terms say the provider is not obliged to remove installed cabling or restore masonry or paint after cancellation. Loan equipment must be returned. Moving is conditional on technical availability. A mesh system, carefully placed access point or building permission may have to be revisited with the next provider. Automatic-debit and bank promotions can disappear on exit.

For a business, the list grows. Public IPv4 addresses may change. Firewall allowlists, VPN peers, remote cameras, point-of-sale systems, domain records and supplier portals may encode those addresses. A provider-managed LAN can contain VLANs, credentials and undocumented dependencies. A multi-site design can make every branch depend on the same operator. A fixed installation may involve landlord, park or building approvals that cannot be recreated quickly.

Fiberty can reduce some switching friction by supplying both access and internal networking, but that is also how dependency deepens. The cure is not to reject an integrated offer; it is to insist on portability:

  • customer-owned domain names and administrative accounts;
  • exportable configurations and diagrams;
  • a clear list of loaned and customer-owned equipment;
  • documented address dependencies;
  • contract language for transition assistance;
  • overlap time for a replacement circuit;
  • and a tested cellular, satellite or second-fixed-access fallback.

The online cancellation form is a positive piece of operational plumbing: it lets the customer request termination and select a reason. The terms likewise say cancellation may be requested at any time and takes effect immediately, subject to identity checks and equipment return. A prudent customer should preserve the confirmation, arrange return evidence and reconcile the final invoice.

Competition is local, physical and increasingly cross-technology

Fiberty does not compete against a single national benchmark. Bariloche’s access market is segmented by street, building, terrain and technology. An independent 2026 provider overview describes more than ten operators and characterises fibre coverage as address-dependent; because it is a commercial comparison site and some displayed prices are evidently fragile, it is useful as a market map rather than an authority on current tariffs.

Official competitor pages make the comparison more concrete. BariNet’s plan page advertises 500-megabit residential fibre and 300, 500 and 600-megabit business tiers, with different support targets; its coverage page distinguishes fibre and wireless areas on a map. AVC’s internet page advertises fibre tiers up to 540 megabits, automatic-debit incentives and a discount for taking television and internet together. These are competitor claims, not independent performance tests.

Fiberty’s differentiation is visible. It stresses buried rather than aerial plant, makes 100/300/500 choices easy to understand, offers mesh, markets a local human team, and uses automatic debit and satellite-equipment exchange to reduce acquisition cost. For enterprises it moves above commodity access into design, security and managed infrastructure.

Its vulnerabilities are equally visible. Competitors publish coverage maps or more explicit business tiers. Fiberty’s retail pages omit upload rates and quantitative service guarantees. Its own ASN is not publicly active. Its enterprise claims are broad but unpriced and unmeasured. And satellite growth gives customers outside the fibre footprint another route to adequate speed without waiting for a trench.

The contest will not be decided by the largest headline number. It will be decided premise by premise:

  • Who can install now?
  • Who controls the last drop and the upstream fault?
  • Who delivers the promised rate at busy hour over a wire?
  • Who restores service on a snowy weekend?
  • Who offers a truly separate backup?
  • Who keeps the quote, support channels and regulatory label current?

For a local ISP, those questions favour density and reputation. Each connected building lowers the marginal cost of the next subscriber. Each well-run outage builds trust; each confused escalation destroys it. PITBA can become a reference account cluster if performance matches the enterprise pitch. It can also become a concentration point if many businesses share one route without knowing it.

A procurement test designed for Fiberty, not for an abstract ISP

The qualification question for Fiberty is simple to state and hard to fake: what separates licence permission from deployed access, active capacity and accountable day-to-day operations? A buyer can answer it through an ordered test.

1. Reconfirm identity and authority

Put the full legal provider name, CUIT and Resolution 2101/2021 in the proposal and invoice. Confirm that the party taking payment, owning the service obligation and handling customer data is the same party. If an installer, portal operator, wholesale carrier or consultant is different, name the role without treating it as the contracted ISP. This preserves the identity boundary already established by the legal, contractual and number-resource records.

2. Obtain an address-specific feasibility document

Require more than “coverage available.” The document should identify building entrance, equipment room, route to the unit, expected installation work, permissions, customer and provider equipment, one-off charges and target date. At PITBA or another managed site, confirm the park’s access rules and who controls each duct segment. Ask whether the quoted route is already lit or depends on new construction.

3. Freeze the service label and technical appendix

Record download and upload rates, committed rate, burst policy, aggregation, data limits, IP addressing, IPv6, translation, router make and specification, Ethernet hand-off and traffic-management rules. For enterprise service, define “dedicated,” “symmetric,” “guaranteed” and “redundant” in measurable language. Attach the regulatory label referenced in the contract.

4. Run a wired acceptance trial

Test by Ethernet at the demarcation point, not only through Wi-Fi. Run repeated measurements in quiet and peak periods to local, Buenos Aires and international destinations chosen in advance. Capture download, upload, round-trip delay, jitter, loss, route changes and application-level behaviour. A public speed-test result is one sample; the acceptance record should be reproducible and tied to the circuit.

5. Make the route visible

Ask which address block the customer will use, which autonomous system originates it, and which upstreams carry it. If the answer is not AS273846, that is not automatically a problem; document the actual provider chain. If AS273846 is to be activated, verify the announcement, valid route authorisation, contact records and failover. Ask whether any traffic is exchanged at IXP Bariloche and what happens when that exchange or a transit path fails.

6. Test physical diversity

Trace both primary and backup services from the premises to the first genuinely independent aggregation point. Two fibres in one duct fail together. Two retail brands on one wholesale carrier may fail together. A satellite or mobile backup may offer stronger physical diversity than a second buried fibre, though it introduces its own power, weather, latency, capacity and equipment trade-offs.

7. Exercise support before signing

Open a non-urgent ticket through each promised channel. Record acknowledgement, ticket number, ownership and escalation. For a business plan, call outside retail hours and confirm the 24/7 path. Ask who can dispatch, who can change routing, who holds spares and who can approve service credits. Put names or roles—not merely “support”—into the escalation appendix.

8. Rehearse a failure

Disconnect the primary hand-off during an agreed test. Verify that the backup route takes over, critical applications reconnect, monitoring alarms, and staff know what to do. Restore the primary and confirm stable return. Repeat after a configuration change. Business continuity that has never been exercised is an aspiration.

9. Audit the portal and data path

Verify encrypted portal access, certificate behaviour, password reset, multi-factor options, role separation, session expiry, data retention and incident contact. Decide which information may be sent by email or messaging. Obtain the privacy notice and list any subprocessors that handle billing, support or network monitoring.

10. Price the whole lifecycle

Use a twelve- or thirty-six-month worksheet including installation, unusual civil work, router or mesh, public addresses, managed equipment, taxes, automatic-debit conditions, price-review terms, backup access, relocation and exit. Treat promotions whose validity has passed as unavailable until reissued in writing. Specify equipment return and transition assistance.

This test is deliberately harder than asking whether the fibre is “real.” It distinguishes the right to sell from the ability to install, the ability to install from the capacity to carry peak traffic, and capacity from the organisation required to restore it.

The watchpoints that will show whether Fiberty is maturing

Several public signals can sharpen the picture over the next year.

First, routing activation. The appearance of the company’s IPv6 block behind AS273846, accompanied by a valid route authorisation and credible upstream diversity, would demonstrate that the number resources have moved from inventory into operation. It would also give customers a clearer public view of route control.

Second, local interconnection. A place in the published IXP Bariloche participant list, or a clear statement of another local-peering arrangement, would make wholesale strategy more legible. Non-membership would not condemn the network, but a documented answer would reduce uncertainty about local traffic and transit dependence.

Third, offer hygiene. Retail prices, promotional dates, support hours and contact routes should agree across the home, plan, contract and cancellation surfaces. A promotion still shown after its stated end date weakens procurement confidence even if the service itself is sound.

Fourth, quantitative enterprise terms. A standard example SLA—availability, acknowledgement, restoration, maintenance notice and credit—would turn the corporate page from a list of capabilities into a benchmark. Customisation can then improve on a visible floor.

Fifth, coverage evidence. A maintained map or address checker that distinguishes serviceable, construction-required and planned premises would help subscribers understand the difference between market presence and immediate access. PITBA is a strong reference point; more independently verifiable sites would deepen the deployment record.

Sixth, incident transparency. A small status page and concise post-incident notes would let customers evaluate operations without exposing sensitive topology. Even a monthly availability statement would be more informative than an absolute promise of no interruptions.

Seventh, customer-edge security. A named privacy notice, secure branded portal, published abuse and security contacts, and clear routing-security posture would align the management surface with the enterprise security pitch.

None of these signals substitutes for a circuit test. Together, they would make Fiberty easier to underwrite as more than a local access brand.

Where the evidence finally lands

Fiberty passes the first and most important identity test. It is not a free-floating brand attached to an evocative name. The exact company was formed for telecoms, licensed by the national regulator, named in its own customer contract, associated with the Fiberty mark, assigned internet-number resources and recognised by Río Negro as the fibre provider at a defined productive and technology park.

It also passes a basic continuity test. The public record spans nearly six years, from incorporation to a live 2026 product and enterprise offer. The address-specific order process, loaned equipment, customer portal, support channels and PITBA listing describe an operator doing the mundane work of access service, not an artificial-intelligence laboratory living only in a registry.

The evidence stops short of proving the most consequential performance claims. The licence does not reveal plant. The park listing does not reveal capacity. A 500-megabit label does not reveal upload speed or aggregation. An allocated ASN does not reveal a live route; at the cut-off, it was quiet. “Redundancy” does not reveal physical diversity. “24/7” does not reveal a tested escalation. “Buried” does not reveal power autonomy or repair time.

That boundary is not a verdict against Fiberty. It is the commercial reality of buying from a regional ISP whose value lies close to the ground. The strongest case for the company is local: a buried access proposition, a tangible industrial-park footprint, an address-aware build process and a support structure that promises humans rather than an anonymous national queue. The risk is also local: a small number of routes, powered sites, technicians and upstream relationships can concentrate failure even when the licence is national and the plan speed is large.

For households, the decision can remain practical: verify the address, obtain a current price, test by wire, check Wi-Fi separately and understand cancellation and equipment return. For SMEs, the bar should be higher. Demand the route, the hand-off, the capacity commitment, the restoration clock, the security owner and the failure rehearsal. If Fiberty can supply those, its unusual legal name becomes irrelevant and its local operating discipline becomes the product. If it cannot, neither Resolution 2101 nor AS273846 can fill the gap.