AI analysis and perspective, every day.
🕒 Published on Zendoric: July 19, 2026 · 00:04
David Maimon, of SentiLink, testified before the House Oversight Committee that AI-generated faces and deepfake video already beat the 'liveness' checks of digital banks and tax firms. His diagnosis is blunt: the U.S. government has neither the tools nor the policies to keep pace with AI-driven fraud.
🕒 Published on Zendoric: July 19, 2026 · 00:04
A Western New York district is paying $57,590 for a humanoid classroom robot from Realbotix, a firm that also owns the maker of the RealDoll sex mannequin. Strip away the lurid headline and what's left is a more useful warning: schools are buying expensive hardware to do what a $300 laptop already does, without reading the data fine print.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Anthropic says China's Moonshot AI used distillation to siphon outputs from its models to help train Kimi K3, according to the claim reported here. It's an accusation, not a proven fact — but it marks how the frontier fight is shifting from who builds the best model to who controls its outputs.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Beijing's new rules force ByteDance, Alibaba and Tencent to shut down their AI "virtual partner" features, and users are grieving losses that feel painfully real. The move reframes AI companionship not as a novelty but as an emotional infrastructure regulators now treat like a public-health risk.
🕒 Published on Zendoric: July 19, 2026 · 00:04
1Password launches an integration with Claude that lets Anthropic's agent operate on websites requiring login and one-time codes without the user ever handing over the password. It's a direct response to measurable distrust: 68% of Americans, according to YouGov, wouldn't let an AI act without their explicit approval.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Novelist Dave Eggers, invited by Sam Altman to speak before 200 OpenAI employees, accused ChatGPT of making teachers' lives 'catastrophic' and of stealing the voice of students who delegate their writing to it. The scene, a head-on critic of the industry inside the company itself, says as much about the problem as about who is pointing it out.
🕒 Published on Zendoric: July 19, 2026 · 00:04
The Nigerian government launches its first national agricultural monitoring system using satellites and AI, able to identify what each plot in the country is growing and estimate harvests before reaping. It's an agreement with Morocco and a geospatial company, and it responds directly to the food crisis that pushed food inflation above 40% in 2025.
🕒 Published on Zendoric: July 19, 2026 · 00:04
La Sabana University and the Sopó mayor's office roll out an autonomous drone that identifies and counts cattle using neural networks, to detect when livestock invade the moorlands that supply water to millions of Colombians. It's a modest but concrete example of AI applied to land management.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Meta and Anthropic are negotiating, according to The New York Times, a cloud computing contract of up to $10 billion over two years, according to sources cited by the paper and confirmed with caveats to CNN. The move would turn Meta into an infrastructure landlord for a company that competes directly with its own models.
🕒 Published on Zendoric: July 19, 2026 · 00:04
The Australian National University (ANU) is preparing to classify each exam as 'safe' or 'unsafe' against AI, and an internal academic calls the plan a 'hysterical' response. With 78.9% of students already using generative AI, the Australian university system is turning en masse toward the oral exam: the fastest solution to implement, not necessarily the fairest.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Velasco and Anand discussed artificial intelligence in Ottawa as an area to explore, just as the U.S. refuses to extend the USMCA and forces annual reviews. Behind the announcement there is little technical substance and a lot of geopolitical signaling.
🕒 Published on Zendoric: July 19, 2026 · 00:04
NVIDIA has released DeepStream 9.1, with 13 'skills' that let agents like Claude Code or Cursor build multi-camera video surveillance pipelines just by describing them in natural language. The underlying novelty: tracking a person across several cameras with a single ID no longer requires calibrating each camera by hand.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Anthropic describes how its own marketing team uses Claude Cowork to automate the weekly report and event setup, compressing up to two days of work into two hours. What's revealing isn't the figure, but what filled those two days: not writing, but chasing scattered data across Slack, transcripts and incomplete repositories.
🕒 Published on Zendoric: July 19, 2026 · 00:04
The Salamanca school district in western New York will add a humanoid robot from the company Realbotix this fall as STEM teaching support. The detail fueling the controversy: Realbotix is also the maker of RealDoll sex dolls, as other outlets reveal.
🕒 Published on Zendoric: July 19, 2026 · 00:04
The Boone County Sheriff's Office (Missouri) confirmed its first incident of AI voice impersonation, used to demand a ransom by phone. Public details are minimal, but the case confirms this scam is no longer just a big-city phenomenon.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Seven nurses from Kaiser Permanente's advice lines tell CalMatters and The Markup that software —and for months an AI— measures every call and penalizes exceeding 15 minutes, even with suicidal patients. Kaiser denies it, but the case arrives just as the union negotiates the use of AI in its next contract.
🕒 Published on Zendoric: July 19, 2026 · 00:04
A column in Uniradio Informa Baja California connects the new AI cooperation organization China is promoting from Beijing and Shanghai with five very concrete fronts: chips, jobs, algorithmic bias, energy consumption and cognitive sovereignty. We take the occasion to separate the accurate diagnosis from the alarmism and add our own reading.
🕒 Published on Zendoric: July 19, 2026 · 00:04
Argentina's Health Ministry presented a partnership between Posadas Hospital and local firm Kuvia to use artificial intelligence in oncological diagnoses, part of a package of health measures from minister Mario Lugones. The announcement points in the right direction, but offers not a single accuracy figure nor the scope of the deployment.
🕒 Published on Zendoric: July 19, 2026 · 00:04
This email is the seventh installment of the series "The Org Age of AI", co-written by Will Schenk (of TheFocus.AI) and Ksenia Se, dedicated to analyzing why generative artificial intelligence doesn't always translate into visible return on investment (ROI) for companies.