Friday, November 20, 2026
HomeStartupA transcription agency utilized by police, prosecutors and tech giants uncovered buyer...

A transcription agency utilized by police, prosecutors and tech giants uncovered buyer information on-line


A transcription firm that caters to legislation enforcement, authorities departments, and Silicon Valley tech giants has resolved a safety lapse after leaving a considerable amount of inside information, together with buyer transcriptions and their supply supplies, uncovered to the net.

The info safety lapse at San Francisco-based Tigerfish uncovered transcriptions of police interviews, body-cam footage, and clients’ uncooked audio and video footage. The uncovered information additionally included invoices of the corporate’s clients, which embrace police departments and prosecutors’ workplaces.

Based on Anurag Sen, a great religion safety researcher who discovered the information, the recordsdata have been situated on an Amazon-hosted (S3) cloud server that was brazenly listed and accessible to anybody through an easy-to-guess net deal with — no passwords wanted.

The safety lapse underscores how easy flaws or misconfigurations may end up in delicate info getting uncovered on-line, particularly as AI makes it simpler to uncover weaknesses that may put firms and their information in danger.

Sen found the uncovered information on-line in October and alerted the corporate to the problem, however didn’t hear again. Sen later supplied particulars to TechCrunch in an effort to flag the incident with the corporate for decision.

Sen isn’t any stranger to TechCrunch. For the previous decade, Sen, 30, has looked for delicate datasets uncovered on-line with the purpose of notifying the businesses and organizations that their info is open to the web. Like many safety researchers, a lot of his efforts are pushed by an obligation to safe the web, and he describes his work as an “unpaid air-traffic controller.”

From his island house in Indonesia, Sen has uncovered delicate U.S. army emails spilling from a compartmentalized Division of Protection cloud server; databases exposing reams of buyer information; the inside workings of a rustic’s nationwide license plate surveillance system; and public programs that inadvertently despatched out streams of SMS codes utilized by individuals around the globe to log into their on-line accounts. 

Given Sen’s propensity for locating typically extremely confidential info on-line, we verified the supply of the leak and the information, and contacted Tigerfish.

Quickly after making contact, Tigerfish secured its uncovered S3 cloud bucket.

“We’ve blocked all public entry to the bucket you recognized, together with listing listings and particular person recordsdata,” Tigerfish president Adam Goldberg stated in an e mail. “We’ve additionally verified that every one of our different S3 buckets are blocked to public entry.”

“We’re persevering with to evaluation and strengthen our storage safety extra broadly,” added Goldberg.

The corporate stated that Sen’s e mail disclosing the information leak was caught in its spam filter.

It’s unclear if the corporate has notified its clients. Tigerfish didn’t reply to a follow-up e mail.

‘Primary human error’

Discoveries of uncovered datasets have consumed a big a part of Sen’s early cybersecurity profession thus far. He scours the net each day utilizing automated scanners and specialist serps to establish databases, cloud storage, or different delicate info that doubtless shouldn’t be left unprotected on-line. 

The trendy net runs on absolute belief, he stated, and operates below the notion that the world’s information is stored safe. However he described how his findings reveal the fact of how “fragile” the web is behind the scenes.

Regardless of the latest concentrate on how AI is accelerating individuals’s capability to find vulnerabilities, Sen stated that most of the information leaks and exposures that he finds come all the way down to fundamental human errors. 

“An engineer spinning up a take a look at cloud database on a Friday, disabling authentication to check an API, and forgetting to close the door earlier than heading into the weekend,” was only one instance, he recounted.

Sen’s findings through the years spotlight the continuing issues brought on by such safety errors, simply as the brand new wave of vibe-coding and AI-generated improvement is creating new safety issues on account of insecure and unchecked code deployed to the net. 

Much like Amazon’s latest efforts to make it tough to make S3 buckets public with out going by clear multi-step warnings, newer vibe-coding platforms like Lovable and Supabase have additionally made modifications to their platforms to forestall consumer errors.

“Everybody typically fears shadowy hackers,” Sen stated. However he argues that the true disaster is leaving delicate programs and databases “broad open to the general public web with no password.”

Once you buy by hyperlinks in our articles, we could earn a small fee. This doesn’t have an effect on our editorial independence.

RELATED ARTICLES

Most Popular

Recent Comments