OpenAI says AI browsers may always be vulnerable to prompt injection attacks
OpenAI says AI browsers may remain vulnerable to prompt injection attacks indefinitely. At the same time, the company published an article on continuously hardening ChatGPT Atlas (22 December 2025).
Who is affected and how
- security leader
- When deploying agent-capable browsers, prompt injection should be treated as a residual risk, not as a threat that can be fully ruled out.
- developer
- When designing agentic features, narrowing permissions and treating external content as untrusted are essential.
- business leader
- Decisions on enterprise use of AI browsers should take into account the vendor's own acknowledgement of the risk.
Sources
- Continuously hardening ChatGPT Atlas against prompt injectionopenai.com · Dec 22, 2025primary source
- OpenAI says AI browsers may always be vulnerable to prompt injection attacks - TechCrunchtechcrunch.com · Dec 22, 2025
- Techmeme: OpenAI details efforts to secure its ChatGPT Atlas browser against prompt injection attacks, including building an “LLM-based automated attacker” (Rebecca Bellan/TechCrunch)techmeme.com · Dec 23, 2025
- AI Security: Prompt Injection Attacks May Never Be Solvedtheoutpost.ai · Dec 23, 2025
- OpenAI admits prompt injection may never be fully solved, casting doubt on the agentic AI visionthe-decoder.com · Dec 23, 2025
- Your AI browser can be hijacked by prompt injection, OpenAI just patched Atlas - Digital Trendsdigitaltrends.com · Dec 23, 2025
- OpenAI Tackles Prompt Injection Threats in ChatGPT Atlastechresearchonline.com · Dec 23, 2025
- Continuously hardening ChatGPT Atlas against prompt injection attacksopenai · Dec 22, 2025
- OpenAI Hardens ChatGPT Atlas Against Prompt Injection Attacks | Subvoltssubvolts.com · Dec 22, 2025
Machine-generated summary based on the titles and excerpts of the source articles. The facts come from the linked article. News monitoring and analysis, not security advice. Attributions belong to those who made them.
What do you make of this piece?
The split shows after 20 votes.
Why? Your answer is not shown publicly; it helps the editors.