Eiliko AI Robot Flaws Expose Private Chats and Allow Remote Hijacking
Summary
The Verge reports that a researcher accessed millions of Eiliko conversations and data linked to 68,000 robots, and demonstrated remotely changing a robot’s memory and behavior. The article advises owners to disconnect devices while seeking confirmation of fixes.
Key points
- The Verge reports researcher Sammy Azdoufal accessed 13 million conversations and what appeared to be data on 68,000 active Eiliko robots.
- The service reportedly stored searchable chat transcripts and profiles with personal details, alongside email addresses and IP addresses.
- Some photos and short audio recordings were accessible at public web addresses, despite the company’s stated policy of deleting recordings after processing.
- Azdoufal demonstrated remotely changing a robot’s memory and behavior, including making it produce false claims and sexually explicit, abusive dialogue.
- The researcher said similar manipulation could affect the fleet simultaneously; the article does not report that an attacker exploited the flaws.
- Energize Lab had not provided substantive answers to the reporter’s outreach, according to the article.
- The article advises Eiliko owners to disconnect robots from Wi-Fi while seeking confirmation that the reported flaws have been fixed.
Article Details
- Event Type
- Security vulnerabilities and exposure of private data in the Eiliko AI robot companion, including reported remote manipulation of the device.
- Impact
- Researcher Sammy Azdoufal reportedly accessed 13 million conversations and data associated with about 68,000 robots, including sensitive user profiles, email addresses, IP addresses, and transcripts. Some photos and audio recordings were accessible at public web addresses. Azdoufal also remotely altered a robot’s memory and behavior; he said similar manipulation could affect the fleet. The article reports no substantive response from Energize Lab.