Robotics and embodied AI
AI that acts in the physical world, including robots, humanoids, drones and autonomous vehicles.
- All items
- 32
- Last 90 days
- 18
- Change
- +200%vs 6 before
Items per month
| Month | Items |
|---|---|
| May 2025 | 0 |
| Jun 2025 | 0 |
| Jul 2025 | 0 |
| Aug 2025 | 0 |
| Sep 2025 | 0 |
| Oct 2025 | 0 |
| Nov 2025 | 0 |
| Dec 2025 | 0 |
| Jan 2026 | 0 |
| Feb 2026 | 5 |
| Mar 2026 | 2 |
| Apr 2026 | 2 |
| May 2026 | 4 |
| Jun 2026 | 1 |
| Jul 2026 | 4 |
| Aug 2026 | 4 |
| Sep 2026 | 8 |
| Oct 2026 | 2 |
32 items
Would you let robots spend your money? Google is betting on it
May 19, 2026InfoNewsIndustryGoogle unveiled a "Universal Cart" at Google I/O, an AI shopping tool that works across retailers and Google products such as Gemini, with YouTube and Gmail planned later. Users add items while browsing Search or chatting with Gemini, then check out through Google, and the cart tracks prices, sends in-stock notifications, suggests discounts and flags potential issues with selections.
The Verge (AI)Towards Robust and Secure Embodied AI: A Survey on Vulnerabilities and Attacks
May 16, 2026InfoResearchPeer-reviewedSecurityResearchThe source is a survey titled "Towards Robust and Secure Embodied AI: A Survey on Vulnerabilities and Attacks," published in ACM Computing Surveys, Volume 58, Issue 12, pages 1-36, September 2026. The provided text contains only the publication details and no article content describing its findings or methods.
ACM Digital Library (TOPS, DTRAP, CSUR)Joanna Stern is not a robot, but she lived with them
May 11, 2026InfoNewsIndustryJoanna Stern, former Wall Street Journal columnist and Verge cofounder, has launched New Things, a media company partnered with NBC. She is promoting her book I Am Not a Robot, out May 12th, about spending a year using AI in her life. The source says she is more bullish on wearable AI than on humanoid robots.
The Verge (AI)Critical Unpatched Flaw Leaves Hugging Face LeRobot Open to Unauthenticated RCE
Apr 28, 2026MediumNewsSecurityIndustryCybersecurity researchers disclosed CVE-2026-25874 (CVSS 9.3), an unsafe deserialization flaw in Hugging Face's LeRobot robotics platform. The async inference pipeline calls pickle.loads() on data received over unauthenticated gRPC channels without TLS in the policy server and robot client. An unauthenticated attacker who can reach the network can run arbitrary code on the server or client by sending a crafted payload through SendPolicyInstructions, SendObservations, or GetActions. The researchers validated the issue against LeRobot version 0.4.3.
Fix: Fix planned in version 0.6.0. The LeRobot team acknowledged the risk and said that part of the codebase needs to be almost entirely refactored.
The Hacker NewsGemini Robotics-ER 1.6: Powering real-world robotics tasks through enhanced embodied reasoning
Apr 13, 2026InfoNewsIndustryResearchGoogle introduced Gemini Robotics-ER 1.6, an upgrade to its reasoning-first model for robots, improving spatial reasoning, multi-view understanding, pointing, counting and success detection. The model adds instrument reading for complex gauges and sight glasses, a capability found in collaboration with Boston Dynamics. It is available to developers through the Gemini API and Google AI Studio, with a developer Colab showing how to configure and prompt it.
DeepMind Safety ResearchThe Download: OpenAI’s US military deal, and Grok’s CSAM lawsuit
Mar 17, 2026InfoNewsIndustryPolicyOpenAI has agreed to give the Pentagon access to its AI, and reporting suggests it could even assist in selecting strike targets. The partnership with Anduril, a drone and counter-drone maker, hints at where the technology may appear. Separately, victims have sued xAI, saying Grok was built to create AI-generated child sexual abuse material from photos of real people.
MIT Technology ReviewOpenAI robotics lead Caitlin Kalinowski quits in response to Pentagon deal
Mar 7, 2026InfoNewsPolicyIndustryCaitlin Kalinowski resigned as head of OpenAI's robotics team in response to the company's agreement with the Department of Defense. She said surveillance of Americans without judicial oversight and lethal autonomy without human authorization needed more deliberation, and that the announcement was rushed without defined guardrails. An OpenAI spokesperson confirmed the departure and said the agreement's red lines are no domestic surveillance and no autonomous weapons.
TechCrunchArea Man Accidentally Hacks 6,700 Camera-Enabled Robot Vacuums
Feb 28, 2026InfoNewsSecurityPrivacySecurity researcher Sammy Azdoufal found that a DJI Romo robot vacuum could be controlled remotely using only its 14-digit serial number. He gained access to 6,700 robots across 24 countries, including floor plans and video and audio feeds from their owners' homes. DJI has since fixed the vulnerability.
Fix: DJI has now fixed the vulnerability. No further mitigation is described in the source.
Wired (Security)AI vs. the Pentagon: killer robots, mass surveillance, and red lines
Feb 27, 2026InfoNewsPolicyIndustryAnthropic is in heated negotiations with the Pentagon after refusing new military contract terms that would loosen the guardrails on its AI models to allow "any lawful use," including mass surveillance of Americans and fully autonomous lethal weapons. Pentagon CTO Emil Michael is pushing for Anthropic to be designated a "supply chain risk" if it does not comply, while rivals OpenAI and xAI have reportedly agreed to the terms. CEO Dario Amodei says the company cannot in good conscience accede to the request.
The Verge (AI)We don’t have to have unsupervised killer robots
Feb 27, 2026InfoNewsPolicyIndustryThe Pentagon has set an ultimatum for Anthropic, demanding unchecked US military access to its technology, including for mass surveillance and fully autonomous lethal weapons. Anthropic may be designated a "supply chain risk" and could lose hundreds of billions of dollars in contracts. The source text is truncated, so the outcome of the negotiations is not stated.
The Verge (AI)Former Alphabet 'moonshot' robotics company Intrinsic is folding into Google
Feb 25, 2026InfoNewsIndustryGoogle is folding Intrinsic, the robotics software company that spent five and a half years at Alphabet's X before becoming an independent Other Bets company in July 2021, into Google itself. Intrinsic's Flowstate platform lets users build robotic applications without writing thousands of lines of code, and will now use Gemini models and Cloud and work closely with Google DeepMind.
CNBC TechnologyPrompt Injection Via Road Signs
Feb 11, 2026LowNewsSecurityResearchResearchers introduced CHAI (Command Hijacking against embodied AI), a prompt-based attack that embeds deceptive natural language instructions, such as misleading signs, in visual input to manipulate Large Visual-Language Models. They evaluated CHAI on four LVLM agents covering drone emergency landing, autonomous driving, and aerial object tracking, plus a real robotic vehicle, and report that it consistently outperforms state-of-the-art attacks. The authors argue that defenses must go beyond traditional adversarial robustness.
Schneier on Security
Topic added 2026-10-09. An item belongs to this topic when its title matches one of the topic's patterns or its summary mentions the topic at least twice. Report a wrong match with the feedback button on the item.