A Porsche Certified Pre-Owned car can lose full warranty protection over tampering that happened years before you bought it. The dealer that certified it evidently couldn’t detect it, and no dealer can undo it.

Porsche’s Certified Pre-Owned program vouches for something its own dealers evidently can’t inspect. I learned that with my own car. In March, Porsche AG placed a warranty block on my 2020 911 Carrera 4S over a DME modification made on July 20, 2021, at 9,096 miles. That was four years and two CPO sales before I bought it. Two certification inspections missed it. A weeklong diagnostic ordered by the arbitrator found nothing the modification had caused. The block remains anyway.

The block doesn’t void my coverage, which runs through July 11, 2028. Every claim now has to be reviewed by Porsche AG, the same body that has already said it can’t rule out damage it can’t find. The buyer, that’s me, who had no way to know about the modification and no way to disprove its effects, now carries the burden of proof.

The facts are not in dispute. I bought the car from Porsche South Austin on July 11, 2025, at 17,300 miles, with CPO coverage plus an extra year. On July 21, two days before I took delivery, the selling dealer logged and repaired “multiple data bus faults.” In September 2025 I took the car to my first Maverick PCA HPDE at Eagles Canyon Raceway. Porsche McKinney performed the required tech inspection and passed it. Near the end of the day, on the hard left at Turn 15, the car went into limp mode. A restart cleared it. Two laps later, on another hard left near the crossover, it happened again. I parked it. Porsche McKinney kept the car from September 24 to 30. Their repair order shows an undocumented DME fault, no aftermarket parts and no repair.

One disclosure, because Porsche will raise it if I don’t. On February 18, 2026, I had an independent Porsche specialist activate the matrix headlights, a factory option already listed on the car’s window sticker. That was five months after the first limp mode episode.

In March 2026 I went back to Eagles Canyon with the Lizard. On the first warm-up lap, exiting a hard left under throttle, the car went into limp mode again. I sent my scan data to Porsche McKinney on March 21 and delivered the car on March 23. In his request to PCNA for help, their technician wrote that “the vehicle appears to be completely stock.” On April 6, PCNA told the dealer that software manipulation had been detected and that a warranty block had been set. I disclosed the headlight coding the next day and had it reverted at my own expense on April 8. On April 23, Porsche AG confirmed that factory DME software was installed. The car had been out of service 38 days. On May 21, PAG refused to lift the block because “it is unclear whether the modifications made over the past few years have caused any pre-existing damage to the drivetrain.”

Over the past few years?  I had owned the car for ten months. I didn’t learn what that phrase meant until the arbitration hearing on July 28, when PCNA’s witness read the 2021 date into the record.

I didn’t go to arbitration to get Porsche to buy the car back, and I didn’t ask for money. I love this car. Outside of Turn 15 at ECR, it’s perfect. I asked for three things: lift the block, handle future claims through the normal process, and diagnose the limp mode under the conditions that produce it, with a technician riding along and PIWIS connected to the OBD port.

PCNA argued that a modified DME breaks the chain of custody. It also cited an August 2025 visit to an independent shop on the Carfax as a possible cause. That visit was a fire extinguisher under the front passenger seat installed by Blair Automotive. The arbitrator ordered a full diagnostic at PCNA’s expense. I delivered the car on August 24. Technicians from Porsche McKinney, PCNA’s Field Technical Manager and Porsche AG spent a week, giving it what I euphemistically called a colonoscopy. On September 4, PCNA conceded in writing that its inspection “did not identify evidence directly linking the reported fault codes to software modifications.” The findings pointed to suggested maintenance: spark plugs and carbon buildup. I paid for the plugs and an oil change myself. Nobody has reimbursed me. PCNA covered the engine’s carbon cleaning.

At 10:33 a.m. on September 8, PCNA recommended replacing the DME to eliminate what it called “potential diagnostic uncertainty.” Its counsel also conceded that a new DME wouldn’t allow PCNA to rule out later modifications. At 12:12 p.m., PCNA said it would not pay for one. So the manufacturer recommended a remedy that doesn’t answer its own concern, then declined to provide it.

On September 22 the arbitrator closed the case without further action. The car is not currently in limp mode, so there was nothing left to order. The decision did record two points worth repeating. The warranty says parts that fail during driving events “may” not be covered, which the arbitrator read as discretion, so an HPDE does not void coverage on its face. And if PCNA can tell whether a fault comes from a modification, it can tell whether modifications existed before my first complaint. I rejected the decision on September 28. Accepting it would have bound Porsche to do nothing. Rejecting it preserves my remaining remedies. Litigation is unlikely, because its cost would probably exceed anything it could recover.

The standard Porsche applies here can’t be met, in my opinion. PCNA’s position is that it can’t rule out modifications that its own diagnostic software can’t detect. I believe no inspection, by anyone, can satisfy a test defined by what inspection can’t see. A warranty administered under that standard depends on someone’s discretion, and the owner has no way to win it is my read.

The dealers aren’t the villains. Porsche McKinney argued hard for removing the block, and the internal tickets I obtained during arbitration show it. PCNA argued for it in part. PAG would not move, because it doesn’t know what it doesn’t know, and only PAG can lift a block. When I asked Porsche South Austin to recertify the car, it too was a dead end. The general manager told me he had never seen anything like this in more than 20 years at Porsche dealerships. The service manager, with more than 24 years at Porsche, said the same. The car passed two certifications and years of dealer maintenance before the factory found what the dealers couldn’t. I had always heard that Porsche stands behind its dealers. That isn’t what I observed.

Porsche can fix this without a courtroom. It should run the same DME integrity check that flagged my car before any car is certified, not years afterward. When a block rests entirely on modifications made before a Porsche-certified sale, Porsche should honor the certification it sold. And it should tell owners what evidence would actually lift a block, because a standard nobody can meet isn’t a standard at all in my opinion.

Until Porsche does those things, my rule for buying CPO is simple. I will not buy another CPO car or pay a premium associated with a CPO car without written confirmation from the general manager and the service advisor that they asked PCNA and Porsche AG to verify the DME shows no tampering or modification. I don’t know whether that would hold up if a car were later blocked for something that happened before the sale. But, it’s more protection than the certificate gave me.

If you’re a consumer advocate and have seen this before, I’d like to hear from you about what I could’ve done differently or should do in the future.