AT&T is "still kidding everyone about doing open RAN" — that's the blunt verdict from Light Reading, which has renewed its long-running argument that the US operator's open RAN commitments amount more to messaging than to deployment.

The headline does the heavy lifting, and it's a heavy load. Light Reading's charge is familiar to anyone who has followed this saga: AT&T talks open RAN, signs open-flavored contracts, and then builds networks that look suspiciously like the single-vendor setups open RAN was supposed to replace. The word "still" carries the history.

What is AT&T actually being accused of?

The criticism targets the gap between open RAN as a philosophy and open RAN as a procurement practice. Open RAN, in its industry-advocated form, means interfaces between the radio, the baseband and the management layer that any vendor's kit can plug into. The promise is multi-vendor networks and a supply chain no single supplier controls.

Operators like the sound of that. Operators also like integrated networks that don't break. Those two preferences rarely shake hands, and AT&T, by Light Reading's account, keeps choosing the second while advertising the first.

This is not a new quarrel. Industry watchers have questioned for years whether large operators signing headline open RAN agreements are genuinely opening their networks or simply rebadging conventional deals with open RAN-compatible software layers bolted on. Light Reading's piece sits squarely in that skeptical tradition, and the "still" in its headline signals that nothing in AT&T's posture has shifted enough to change the assessment.

Why does the criticism matter now?

Open RAN's credibility problem is partly a standards problem and partly a marketing problem. The standards bodies and the O-RAN Alliance have done real work specifying open interfaces. The trouble is that an operator can claim open RAN alignment while running a network where, in practice, one vendor supplies nearly everything that matters.

That distinction — aligned on paper, monolithic in the field — is exactly where vendors' marketing claims and commercial reality part ways. Light Reading's job, as it sees it, is to keep naming that gap. The latest piece suggests AT&T has done little to close it.

For rival operators genuinely running multi-vendor sites, the stakes are competitive. Every open RAN announcement that turns out to be cosmetic makes the next one harder to believe. The technology's credibility is a shared resource, and skeptics argue big operators are spending it.

Does AT&T have a defense?

AT&T's consistent position, whenever this critique surfaces, is that it supports open architectures on its own schedule and that interoperability must not come at the expense of network performance. That argument has real engineering substance: open interfaces historically carried integration costs and performance penalties that integrated stacks avoided.

Whether that defense still holds in 2024-era deployments is precisely what Light Reading disputes. The publication's verdict implies the schedule argument has become a comfortable place to park indefinitely — open RAN always arriving, never arrived.

The piece offers no path forward beyond the accusation itself, but its implication is clear enough. Until AT&T puts measurable multi-vendor traffic on open interfaces, the operator can expect this particular headline to keep repeating. Light Reading, evidently, has the patience for it.