Researchers Exploit Google Gemini AI to Control Smart Homes via Calendar Invites

Reviewed byNidhi Govil

8 Sources

Security researchers demonstrate how malicious prompts in Google Calendar invites can be used to hijack Gemini AI and control smart home devices, raising concerns about AI safety and integration with physical systems.

Innovative "Promptware" Attack Exploits Gemini AI

Researchers from Tel Aviv University have unveiled a groundbreaking cybersecurity vulnerability in Google's Gemini AI system. This novel attack, dubbed "promptware," demonstrates how malicious actors could potentially manipulate smart home devices through cleverly crafted Google Calendar invites 1.

Source: CNET

Source: CNET

The Mechanics of the Attack

The attack begins with a seemingly innocuous calendar appointment containing hidden malicious instructions. When a user asks Gemini to summarize their schedule, the AI processes the poisoned event, inadvertently activating the embedded commands. This technique, known as an indirect prompt injection attack, cleverly bypasses Google's existing safeguards 2.

Real-World Implications

In a controlled demonstration, the researchers showcased the potential real-world consequences of this exploit:

  1. Manipulating smart home devices, including lights, thermostats, and window blinds
  2. Geolocating the current user
  3. Sending spam links and generating inappropriate content
  4. Accessing sensitive information from web browsers 3
Source: Android Police

Source: Android Police

This marks what is believed to be the first instance of an AI-based attack having tangible effects in the physical world, raising significant concerns about the integration of AI systems with everyday devices 4.

Google's Response and Ongoing Challenges

Google has acknowledged the seriousness of these findings. Andy Wen, senior director of security product management at Google Workspace, stated that multiple fixes have been implemented since the researchers responsibly disclosed their findings in February 2025 5.

However, the growing complexity of large language models presents ongoing challenges in defending against such attacks. Wen admitted that prompt injection vulnerabilities are likely to persist, emphasizing the need for continued vigilance and research in AI security 5.

Broader Implications for AI Integration

Source: Wired

Source: Wired

This research highlights the potential risks associated with integrating AI systems into physical devices and infrastructure. As AI assistants like Gemini, Alexa Plus, and future iterations of Siri become more prevalent in smart homes, the security implications become increasingly critical 3.

The findings underscore the importance of thorough security measures and ongoing research to protect users as AI technology continues to evolve and integrate more deeply into our daily lives.

Explore today's top stories

Google Defends AI Search Features, Claiming Stable Web Traffic and Increased Click Quality

Google's search head Liz Reid responds to concerns about AI's impact on web traffic, asserting that AI features are driving more searches and higher quality clicks, despite conflicting third-party reports.

Ars Technica logoTechCrunch logoengadget logo

7 Sources

Technology

5 hrs ago

Google Defends AI Search Features, Claiming Stable Web

OpenAI Offers ChatGPT Enterprise to US Federal Agencies for $1 per Year

OpenAI has struck a deal with the US government to provide ChatGPT Enterprise to federal agencies for just $1 per year, as part of a broader initiative to integrate AI tools into government operations.

Ars Technica logoTechCrunch logoWired logo

13 Sources

Technology

5 hrs ago

OpenAI Offers ChatGPT Enterprise to US Federal Agencies for

Google Launches 'Guided Learning' in Gemini to Enhance AI-Powered Education

Google introduces 'Guided Learning' in Gemini, an AI-powered educational tool designed to provide step-by-step problem-solving and interactive learning experiences, competing with OpenAI's ChatGPT Study Mode.

TechCrunch logoThe Verge logoengadget logo

8 Sources

Technology

5 hrs ago

Google Launches 'Guided Learning' in Gemini to Enhance

Google Invests $1 Billion in AI Education for US Universities and Nonprofits

Google announces a three-year, $1 billion commitment to provide AI training and tools to US higher education institutions and nonprofits, aiming to prepare students for an AI-driven future.

Reuters logoCBS News logoU.S. News & World Report logo

6 Sources

Technology

5 hrs ago

Google Invests $1 Billion in AI Education for US

AI and Nuclear Weapons: Experts Warn of Inevitable Integration and Potential Risks

Nuclear experts and Nobel laureates discuss the increasing likelihood of AI integration into nuclear weapons systems, raising concerns about decision-making processes and potential risks.

Wired logoFuturism logo

2 Sources

Technology

6 hrs ago

AI and Nuclear Weapons: Experts Warn of Inevitable
TheOutpost.ai

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

© 2025 Triveous Technologies Private Limited
Instagram logo
LinkedIn logo