- On 19 March 2026, TIM and Fastweb + Vodafone signed a non-binding agreement to develop and operate up to 6,000 new mobile-tower sites in Italy through an initially equal joint venture.
- The announcement concerns new passive infrastructure, not a sale of existing towers or a completed transaction. No purchase price, enterprise value or committed completion date was disclosed.
The proposal is specific—and still conditional
The joint TIM release and Fastweb + Vodafone release describe a 50/50 vehicle that would build and manage new passive mobile infrastructure over a multi-year programme. “Up to 6,000” is a ceiling for prospective new sites, not a count of towers already built, transferred or financed. The project remains subject to the necessary approvals.
Fastweb + Vodafone is the combined Italian operator within Swisscom after Vodafone Italia was incorporated into Fastweb. The counterparties are therefore TIM and Fastweb + Vodafone, not three independent operators. The old wording obscured both the party structure and the preliminary status.
The operating model matters more than a generic claim about tower value
TIM and Fastweb + Vodafone would be long-term anchor tenants, buying passive-infrastructure services from the venture at market prices. Other telecom operators could obtain access, and the parties may later admit third-party investors to improve the capital structure. This model could pool site-development risk and spread fixed costs, but the announcements do not quantify rent, capital expenditure, financing, tenancy ratios or expected returns.
The project therefore cannot support a transaction valuation or prove that every tower asset produces stable multi-tenant income. Its commercial case depends on which sites are permitted and built, how quickly they gain tenants, the lease terms, construction costs and whether open access attracts genuine third-party demand.
It is separate from the active RAN-sharing project
A January 2026 preliminary agreement covers sharing active radio-access infrastructure in municipalities with fewer than 35,000 inhabitants and envisages about 15,500 sites per operator by the end of 2028. The March tower proposal instead concerns new passive structures. The distinction affects ownership, control, competition and regulatory review.
On 21 April, the Italian Competition Authority opened proceeding I882 into the RAN-sharing arrangement under Article 101 TFEU, with a stated completion deadline of 30 April 2027. That investigation is not an approval or rejection of the separate tower venture, but it shows why the two projects must not be collapsed into one headline.
INWIT contracts are a material dependency
Fastweb + Vodafone gave notice to end its INWIT master service agreement, saying a multi-year migration should preserve continuity through and after March 2028. TIM later notified an August 2030 termination, or alternatively March 2028 depending on the relevant change-of-control question. INWIT disputes TIM's position, says its contract runs to August 2038 and argues that its network is difficult to replicate.
INWIT's first-quarter disclosure records the notices, litigation and its contention that roughly 75% of its infrastructure is not replicable. Those are contested company positions, not settled findings. They make transition timing, site rights, service continuity and legal outcomes central to the tower venture's economics.
What the announcement actually signals
Swisscom's first-quarter report continued to describe the tower project as a strategic initiative and excluded new Italian tower agreements from a guidance footnote. The European Commission's 5G Observatory summary confirms the announced scale and anchor-tenant model. The public record reviewed through 14 July 2026 does not establish that a binding final agreement has closed or that construction has begun.
The useful watchpoints are a binding venture agreement, regulatory decisions, financing and lease terms, a site-by-site construction schedule, third-party tenancy, resolution of the INWIT disputes and evidence that rollout improves coverage without inefficient duplication.

