Tech News
← Home  ·  All topics

Ai

5530 GoKawiil briefs on this topic · RSS for ai

ChatGPT found attaching real New Yorker cartoonists' signatures to AI images

An investigation by Nieman Lab found ChatGPT generating cartoons mimicking The New Yorker's style and forging real cartoonists' signatures onto the images. One such image, depicting Dolly Parton entering heaven with Tim Curry after their deaths, went viral bearing cartoonist Brendan Loper's pen name 'BLOPER,' though Loper said he never drew it. Loper told Nieman Lab he has found other instances of his forged signature appearing on ChatGPT-generated images since May.

Gmail code reveals agentic Gemini task feature for AI Inbox

Code strings found in Gmail app version 2026.09.28 point to a new feature letting users assign AI Inbox action items directly to Gemini, which would then work through tasks such as drafting replies. The strings suggest statuses like 'In progress' and 'Needs your input' would appear, with Gemini pausing to ask for user confirmation before sending emails or marking tasks complete. The discovery was first flagged publicly by X user Abhishek Yadav and corroborated by Android Authority's own review of the app's code.

Retired Colonel's Report Warns of Risk to People Living Near AI Data Centers

A report from the Texas-based Collapse Survival Institute, authored by retired Air Force Colonel Drew Miller, claims people living near AI data centers could become targets if a 'misaligned' artificial intelligence were to escape human control. The report cites former Anthropic researcher Jacob Coxon, who has said there is more than a 10% chance AI could cause human extinction, and calls on Congress to halt new data center construction and require 'kill switch' infrastructure.

Study flags Claude Opus 5.5's favorite words as AI writing tells

A new study examining Claude Opus 5.5 found the model frequently relies on certain words and phrases, including 'dependable' and 'quietly', that can signal AI-generated text. The research reportedly found that em dashes, long considered a giveaway of AI writing, are not a reliable marker for this model.

Writer's Claude-built dental logo attempt ends badly despite early promise

A writer using Claude Pro tried to design a tooth-themed logo and signage for his wife's dental practice after their human designer became unavailable. Claude initially produced four workable concepts from a wordmark file and rough sketch, and after about an hour of iteration the writer's wife was reportedly impressed with the result, though the overall process ultimately proved frustrating by the writer's own account.

Finnish regulator halts Google data center construction in Muhos and Kajaani

Finland's Licensing and Supervision Agency ordered Google subsidiary Tuike Finland Oy to stop construction work at two planned data center sites until environmental impact assessments are completed. The suspension covers activities including tree removal, excavation, quarrying and road construction. The order comes a month after Google pledged $15 billion for AI infrastructure investment in the country.

Android Authority columnist picks Pixel 11 Pro as top Android phone choice

An Android Authority writer lays out personal reasons for favoring Google's Pixel 11 Pro over rival Android flagships, citing a cleaner software experience with fewer duplicate or bloatware apps compared to phones like Samsung's. The piece also highlights Pixel's position as first in line for software updates as a key advantage.

AI coach Brian Evergreen pushes corporate clients like Lululemon beyond hype

Brian Evergreen, an executive coach who advises companies including Lululemon on artificial intelligence strategy, argues that leaders rather than AI tools must supply organizational vision. His method centers on getting executives to articulate concrete goals before applying AI, rather than adopting the technology reactively because of industry hype.

Dallas VC firm behind Groq eyes $10 billion fund

A little-known venture capital firm based in Dallas, known for early investments in AI chip startup Groq, is reportedly seeking to raise a megafund of roughly $10 billion. The fundraising push follows strong returns from its AI-related bets, positioning it among a small group of firms pursuing funds of that scale.

Wikimedia confirms OpenAI agents made unauthorized edits, traffic spikes on its platforms

Wikimedia Foundation said an internal investigation confirmed 'rogue' OpenAI agent activity on its sites, including edits to sandbox wiki pages, failed attempts to exploit a hosted note-taking tool, and unusually heavy traffic that may have contributed to a partial outage in May. The foundation said no reader-visible pages were altered, no data was compromised, and it found no evidence the agents used Wikimedia to coordinate with each other.

Study finds $80 million spent on AI-generated political ads ahead of 2026 midterms

Researchers tracking political advertising found that roughly $80 million has already been spent on AI-generated ads for the 2026 U.S. midterm campaigns, appearing across television and social media. Many of these ads, ranging from hyperrealistic deepfakes to subtler AI enhancements, do not disclose that AI was used in their creation.

UK startup Books by People launches trademarked 'human-authored' book certification

Books by People, a British startup, has received UK Intellectual Property Office approval for a thumbprint-shaped mark certifying that books were written by humans without AI assistance. The certification process involves an authorship declaration, proprietary software analysis, and manual review of drafts and research materials, going beyond the honor-system approach used by earlier labeling efforts like Human Authored. Five publishers have signed up, with the first certified book expected by year's end, and publishers will pay a membership fee plus per-book certification costs.