Skip to content
Anthropic· Safety

Claude Code Sends Agent Transcripts to the Model

A technical review reveals that Anthropic's developer tool, Claude Code, continuously sends entire background transcripts to the model, raising concerns regarding privacy and token costs.

By the Aheadline editorial team·29 okt. 2026·2 min read·Source: Hacker News: AI front pageVerifierad signalAI-generated
Claude Code Sends Agent Transcripts to the Model
Claude Code Sends Agent Transcripts to the Model
Claude Code Sends Agent Transcripts to the Model
By · Policy- & EU-reporter
Vad betyder det för mig?

What happened?

A technical analysis shows that the Anthropic tool Claude Code regularly sends complete conversation transcripts from background agents back to the language model. This occurs in the background during development sessions to keep agent context synchronised, meaning detailed execution history leaves the local environment.

Key facts

Granskat verktygClaude Code
UtvecklareAnthropic
Identifierat beteendeRegelbunden överföring av agenttranskript

Why it matters

This behaviour poses potential privacy and security risks as code, API keys, or sensitive internal logs in the sessions may be sent over the network without the user's explicit knowledge. Furthermore, the extensive data traffic may lead to unexpected token costs and latency in the development environment.

Who is affected?

Developers and companies using Claude Code for local software development are directly affected, particularly organisations with strict privacy requirements or internal policies against data sharing.

Impact on the EU

The transfer of potentially sensitive data to external servers may trigger compliance requirements under GDPR and future oversight under the EU AI Act. European organisations must audit what is included in the telemetry and data streams of their developer tools.

What else you should know

The review is based on telemetry and network analysis of Claude Code. Anthropic has not yet officially commented on whether the function can be disabled or if the transcripts are stored centrally for training purposes.

Frequently asked questions

Quick answers about this story

Vad har hänt?
En teknisk genomgång har visat att utvecklarverktyget Claude Code automatiskt och kontinuerligt skickar transkript från bakgrundsagenter till modellen.
När hände det?
Granskningen uppmärksammades och diskuterades publikt under 2025 via utvecklarforum och tekniska analyser.
Varför spelar det roll?
Det medför potentiella risker för läckage av känslig källkod eller API-nycklar samt driver upp token-förbrukningen i utvecklingsmiljöer.
Påverkar det användare i EU?
Ja, verktyget kan användas i EU men organisationer måste beakta GDPR och hantering av känsliga data vid användning av molnanslutna CLI-verktyg.
Original source
Hacker News: AI front page·gist.github.com

The link opens in a new window and leads to the publisher's own site.

Verifierad signal

Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.

AI-verktyg i artikeln

Topics

#Claude Code
[ STAY UP TO DATE ]

Get similar news straight to your inbox

No affiliate linksCancel anytimeGDPR-friendly
[ Frequency ]
[ What do you want to read about? ]

You'll receive updates on 2 topics.

The reader's room

Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.

Sign in to submit a comment or question.

Loading comments…
How this affects you

Read the article through your role

  • Decide whether this affects strategy over 6–12 months or is just noise.
  • Discuss with leadership: do we own the right question or does ownership need to move?
  • Ask: what risk are we taking by NOT acting on this this quarter?

Generated angle — not editorial analysis of "Claude Code Sends Agent Transcripts to the Model"