OpenAI bugging case defendant set for January trial
A Saudi national accused of planting a recording device at OpenAI faces a swift trial following his arrest in October.

The short version
- Alfozan is accused of planting a recording device wrapped in toilet paper inside a television at OpenAI's San Francisco headquarters.
- The defendant faces one count of intercepting oral communications and a maximum of five years in prison if convicted.
- Prosecutors have collected 160 hours of recordings but have not yet provided evidence of espionage or theft of trade secrets.
- The defense plans to use Anthropic's Claude to review the evidence provided by the government.
Why it matters
This case sets a precedent for how corporate espionage and internal security breaches are handled in the AI industry. It also raises unique legal questions about using competing AI models to analyze sensitive evidence during a criminal trial.
The bugging of OpenAI
A federal judge has scheduled a trial for January 20 to address the criminal charges against Abdulrahman Alfozan. Alfozan, a Saudi national, worked as an engineering lead at OpenAI before his arrest in October. He is accused of installing a recording device in a conference room at the company's San Francisco headquarters.
According to the prosecution, OpenAI staff discovered the device in September. The recording gadget was reportedly wrapped in toilet paper and hidden inside the back of a television. Surveillance footage allegedly captured Alfozan looking at the device while it was in the room.
OpenAI fired Alfozan following his arrest. However, the company has stated that it currently lacks proof that he took any customer data or source code. The legal proceedings will determine whether his actions constituted a criminal violation of communication privacy.
Evidence and allegations
The government has seized several items from Alfozan, including a tablet, two laptops, and one other phone. Investigators have been unable to unlock Alfozan's primary phone, but they have successfully obtained 160 hours of recordings from the conference room meetings. While the recording device itself is encrypted, the prosecution noted that data could be recovered if it was backed up to another device.
Despite initial claims that the bugging could pose a danger to U.S. national security, prosecutors have not yet presented evidence of espionage. Alfozan's attorney, David Rizk, argued that the government has been slow to turn over evidence. He specifically noted that the prosecution has not shown that Alfozan engaged in any form of espionage.
Assistant U.S. Attorney Nick Parker told the court that he is not currently suggesting the defendant stole anything. The judge has ordered the government to turn over all FBI reports by November 24. Additionally, the court wants to see any evidence regarding the theft of trade secrets or the defendant acting as an unregistered foreign agent.
A speedy legal timeline
The trial date of January 20 represents an unusually fast turnaround for a federal criminal case. Typically, these cases take much longer to reach a courtroom. The speed of this timeline is largely due to Alfozan's attorney seeking a speedy trial after complaining about the pace of evidence discovery.
Discovery is the legal process where the prosecution must provide the defense with the evidence they intend to use. Judge Vince Chhabria told Rizk that the government must move quickly to provide the defense with the evidence they are entitled to. This ensures that the defense has adequate time to review the materials before the trial begins.
The prosecution has expressed concern regarding how the defense plans to handle the evidence. They are worried about the defense's intent to use an AI tool to review the materials. Because some of the seized evidence may contain proprietary OpenAI data, the government is wary of using a third-party platform to process it.
Using competing AI tools
Alfozan's attorney, David Rizk, told reporters that he intends to use Anthropic's Claude to review the evidence. Anthropic is a direct competitor to OpenAI and was founded by former OpenAI employees. This creates a unique situation where a company's own rival might be used to analyze evidence in a criminal case involving the company's security.
The prosecution expressed a serious concern about using a competing AI tool for this purpose. They highlighted that one of the seized laptops might contain proprietary data belonging to OpenAI. They argued that feeding this information into a competitor's system could lead to further data exposure.
Judge Chhabria told the prosecution that he was less concerned about this than they were. However, he did order the prosecutor to submit a motion by Friday to explain why the defense's use of a competing AI might be a problem. The court will have to balance the defense's right to use technology with the need to protect sensitive corporate secrets.
What to watch
- Whether the defense's use of Claude will be permitted by the judge.
- If prosecutors produce evidence of economic espionage or trade secret theft before January.
- The final verdict on the count of intercepting oral communications.
Try it
Look up the current security protocols for physical offices in tech companies to see how they handle unauthorized recording devices.
Sources and links
- Governmentjustice.gov/usao-ndca/pr/saudi-national-arrested-installing-surreptiti justice.gov
- ReportingCNBC: the article we sourced
First reported by CNBC. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.