Ghosh Warns on AI Security Gaps After OpenAI and Anthropic Breaches
*Ajoy Ghosh, founder of The Cyber Alchemist, addressed rising AI-related cybersecurity threats in a Bloomberg interview following disclosures from OpenAI and Anthropic.*
Ajoy Ghosh, founder and chief information security officer at The Cyber Alchemist, discussed cybersecurity measures companies should adopt after recent hacking incidents at Anthropic and OpenAI. He appeared on Bloomberg's Asia Trade program with Haidi Stroud Watts and Paul Allen on August 3.
The interview focused on practical steps for organizations that rely on large AI models. Ghosh framed the incidents as signals that current defenses around model access, data pipelines, and third-party integrations remain inadequate.
No specific technical details from the breaches were released in the segment. Ghosh instead stressed the need for tighter controls on prompt handling, logging of model queries, and verification of training data sources.
Limited public information
Bloomberg provided only a high-level summary of the conversation. The network did not publish transcripts or additional quotes from Ghosh on the exact solutions he recommended.
The timing aligns with public statements from both OpenAI and Anthropic that they had detected unauthorized access attempts targeting their systems. Neither company has disclosed the scope or impact of those attempts.
Why it matters
Security teams at firms deploying AI tools now face pressure to reassess access policies that were written before models became core infrastructure. Ghosh's appearance underscores that the conversation has moved from theoretical risk to immediate operational response, even if concrete guidance remains sparse.
---
Sources:
No comments yet