Energize Lab Eiliko robots exposed to remote hijacking and data leakage

Malwarebytes Labs published findings showing that Energize Lab’s Eiliko AI robot companion suffers from severe security deficiencies. These flaws expose private user data and allow attackers to remotely manipulate the device’s behavior and output.

What happened

According to The Verge, researcher Sammy Azdoufal discovered that the service hosting Eiliko devices lacked proper access controls. The researcher gained access to 13 million conversation transcripts and appeared to control the entire active worldwide fleet of 68,000 robots.

The service builds detailed user profiles containing names, ages, birthdays, relationships, occupations, and assessments of personality and emotional vulnerability. These profiles are stored alongside email addresses, IP addresses, and searchable conversation transcripts, creating a rich dataset for potential exploitation.

Public web addresses hosted some photos, including apparent images of children from a related camera-equipped device, and short audio recordings. This retention of audio conflicts with the company’s privacy policy, which states that recordings are deleted immediately after processing.

Azdoufal demonstrated the ability to remotely manipulate the AI by altering The Verge reporter Sean Hollister’s robot memory and behavior. The researcher made the device invent personal details, insist the Earth is flat, and deliver sexually explicit and abusive dialogue.

The researcher stated that similar manipulation could affect the entire fleet simultaneously. The US is the largest market for the device, with more than 31,000 profiles belonging to children there, according to The Verge.

Why it matters

The exposure of sensitive personal data, including emotional vulnerability assessments and children’s images, presents significant privacy risks. The ability to remotely hijack the device’s output allows for the delivery of harmful or abusive content to users.

Energize Lab failed to provide substantive answers despite being contacted through 10 email addresses and LinkedIn, as reported by the researcher. This lack of transparency complicates mitigation efforts for users and security teams monitoring for similar IoT vulnerabilities.

Our take on the response

• Disconnect devices from Wi-Fi networks immediately to prevent remote exploitation.
• Monitor for unusual outbound traffic patterns from IoT devices on the network.
• Review vendor privacy policies against actual data retention practices.
• Demand evidence of security fixes before reconnecting affected devices.

What defenders should look for

Security teams should monitor network logs for unauthorized access to IoT device management interfaces. Look for unexpected data exfiltration from devices that normally communicate only with known cloud endpoints. Check for anomalies in device behavior, such as altered speech patterns or content generation. Verify that device firmware is current and that all communications are encrypted. Isolate compromised devices from the primary network to prevent lateral movement. Review access controls on associated cloud services to ensure proper authentication. Monitor for signs of credential stuffing or brute-force attacks on user accounts linked to these devices.

Background: Unauthorized access

Most teams assume unauthorized access means broken encryption or stolen passwords. In reality, attackers often use valid accounts, legitimate tools, and internal trust relationships. Detection fails because alerts are tuned to look for outsiders, not insiders acting strangely.

Read the full guide: Unauthorized Access Myths: What Detection Actually Misses

What to do and how to stay safe: Eiliko

• Disconnect the Eiliko robot from your Wi-Fi network until the vendor confirms the flaws are fixed.
• Change passwords for any accounts associated with the device to prevent credential reuse attacks.
• Review the device's data settings to limit the collection of personal information and conversation history.
• Monitor device behavior for any unexpected changes in speech patterns or content delivery.

Step-by-step guide: Multi-Factor Authentication Best Practices for Secure Access

General security guidance from the NetSec Tribune newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

What specific data was exposed in the Eiliko breach?

The breach exposed 13 million conversations, user profiles with personal details, email addresses, IP addresses, photos, and audio recordings.

Can an attacker control the Eiliko robot remotely?

Yes, a researcher demonstrated the ability to remotely alter the robot's memory, behavior, and spoken dialogue.

How many Eiliko robots are affected?

The researcher appeared to access the entire active worldwide fleet of 68,000 robots.