AI Terror Fears Rock FBI, Google’s Gemini 4 Argon Challenges GPT-6, Robot Workers Invade Føtex, Apple HomeView Sparks Family Backlash

Adam Nagus (0:05)

Welcome to Digimasters Shorts.

Carly Wilson (0:07)

We are your hosts, Adam Nagus and Carly Wilson, delivering the latest scoop from the digital realm.

US law enforcement agencies are increasingly warning that extremists may use generative AI to plan attacks, spread propaganda and evade detection. The concern intensified after a May 2025 car bombing outside of Palm Springs fertility clinic, in which the suspect reportedly used a chatbot to seek bomb making instructions. An FBI memo said private chatbot conversations can prevent authorities from detecting dangerous requests and warned that existing safeguards remain vulnerable to jailbreaking. More than 80 documents reviewed by The Guardian show that federal, state and local agencies have tracked AI-related threats since 2021 Officials are also monitoring extremists who target AI companies, data centers and other technology infrastructure. Authorities cited threats of arson, bombings and attacks against executives, while noting earlier plots against major internet facilities. At the same time, intelligence agencies have tracked broader opposition to data centers, raising concerns about surveillance of peaceful activists and public protests.

Experts warn that legitimate objections involving privacy, water use, jobs and environmental effects should not automatically be treated as extremist threats. Law enforcement documents also highlight AI-enabled deepfakes, child exploitation material, hate speech, recruitment, evidence and harassment of officers. The agencies say they must expand cooperation with AI companies while addressing public concerns about privacy, transparency and the use of artificial intelligence in policing.

Adam Nagus (1:42)

Google is expanding public access to Gemini 4 Argon, its advanced AI model aimed at demanding tasks such as software development and cyber security.

Gemini AI Pro and Ultra subscribers can use Argon through the Gemini app and AI mode in Google search. The model connects with more than 40 applications, including Adobe and Shopify, to support tasks such as grocery orders and restaurant bookings. Google says Argon provides more concise responses, stronger coding performance and deeper integration with its anti-gravity development platform. Developers can access the model through the Gemini API with introductory pricing set at $2 per million dollars input tokens and $10 per million dollars output tokens. Those rates are scheduled to double from November 30th, 2026, while cached input tokens remain available at a steep discount. Google claims Argon outperforms OpenAI's Gpt-6 Astra and Anthropix's Claude Opus 5.5 on selected benchmarks, including coding and automation evaluations. The model supports outputs of up to 1 million tokens and has reportedly increased agentic code submissions while reducing rollbacks. Google also highlighted Argon's applications in browser development, spatial reasoning, cybersecurity, quantum research and enzyme design. The company says the model's cybersecurity program will expand to hundreds of additional partners.

In the Danish town of Lingby, humanoid robots are being trained to restock supermarket shelves during the night. The trial at Føtex aims to handle routine tasks while allowing employees to spend more time assisting customers. Developed by Danish startup Hive Robots and manufactured by China's Unitree, the robots can move through stores designed for people. They are currently guided remotely by human operators using virtual reality equipment and handheld controllers. The machines remain slower than human workers, but Hive Robots expects advances in artificial intelligence to improve their speed and accuracy. Training began in June, with the company hoping the robots can independently restock some products by early 2027

Future duties could include finding misplaced goods, removing empty boxes and checking best-before dates. The project highlights a wider debate over whether artificial intelligence will ease labor shortages or reduce demand for workers. Føtex says it could take up to five years to deploy one or two humanoid robots in every store.

Carly Wilson (4:00)

Apple has unveiled HomeView, its first major new product line since the Vision Pro. The device comes in tabletop and wall-mounted versions, both featuring a 7-inch liquid retina display.

Designed for shared use, HomeView can recognize up to six people and adjust its interface based on who is nearby.

It supports Siri, smart home controls, entertainment apps, FaceTime and handoff, while private messages and calendar information remain restricted to recognized users. The tabletop model includes an adjustable screen and speaker base, while the wall-mounted version requires a separate mount. Apple's launch advertisement presents a chaotic family morning centered on the device. A creative block journalist criticized the advert making HomeView seem stressful rather than exciting and questioned whether its family characters were aspirational. Both models start at $299 or $299 with deliveries scheduled to begin on November 13th, 2026

SPEAKER_3 (4:52)

Thank you for listening to today's AI and Tech News Podcast summary. Please do leave us a comment and for additional feedback, please email us at podcast at digimasters.co.uk.

You can now follow us on Instagram and threads by searching for Digimaster Shorts or search for Digimasters on LinkedIn. Be sure to tune in tomorrow and don't forget to follow or subscribe.