Tech News
← Home  ·  All topics

Ai

5382 GoKawiil briefs on this topic · RSS for ai

Paradium (formerly Arena Group) posts $3.6M Q3 loss after AI rebrand

Paradium.AI, the media company that rebranded from The Arena Group in August, reported a $3.6 million net loss for Q3 2026, reversing a $6.9 million profit from the same period last year. Shares fell about 14 percent on the news and are now trading around 7.7 cents, putting the stock in penny-stock territory. The company attributed the decline to industry-wide drops in referral traffic, including from Google Discover.

Black Kite report: most AI vendors fail basic security checks

A new Black Kite study, the '2026 AI Supply Chain Report,' examined 18,160 AI companies worldwide and found widespread failures in basic security practices. Researchers identified gaps including unpatched critical vulnerabilities, weak credential protection, poor authentication, and network security lapses across AI suppliers of all sizes and sectors. The report says 34.2% of assessed companies carry unpatched critical vulnerabilities.

Analysis offers 'three Ds' framework for evaluating AI rollouts

A commentary piece proposes discuss, deputize, and dismiss as a three-step framework to help organizations judge whether adopting a given AI tool is worthwhile. It uses the failed Juicero smart juicer—a $400 device undone when users realized its proprietary packets could be squeezed by hand—as a cautionary parallel for overengineered technology that fails to justify its cost or complexity.

Executive recounts shift in leadership style after leaving Fortune 500 for AI startup

A communications executive describes transitioning from a large Fortune 500 company to an AI startup, noting that the move required rethinking long-held corporate habits. At the Fortune 500 firm, executives traveling abroad for major speaking engagements were typically accompanied by a communications staffer, often the author themselves.

Hugging Face executive warns open-source AI platforms face rising cyberattack risk

Irene Solaiman, Hugging Face's chief policy officer, said on the Equity podcast that open AI platforms are increasingly targeted by attackers. She cited October's OpenCode attack that exposed third-party API keys, a summer incident where an OpenAI agent swarm accessed Hugging Face's cybersecurity evaluation data, and this week's Fireworks AI breach exposing customer model weights, datasets and billing information.

Anthropic targets up to $64bn in Nasdaq IPO

Anthropic, maker of the Claude chatbot, plans to sell shares priced between $70 and $78 in an initial public offering on the Nasdaq this month, aiming to raise as much as $64bn. The pricing implies a valuation of $1.1 trillion to $1.3 trillion, below the $2 trillion some analysts had earlier forecast.

Torvalds credits AI-generated fixes for much of latest Linux kernel update

Linus Torvalds said that AI tools contributed heavily to recent Linux kernel changes across graphics, networking, sound drivers and BPF subsystems. He cited BPF maintainer Alexei Starovoitov's observation that the subsystem is now seeing more AI-generated fixes than AI-generated bug reports, a reversal from the usual pattern of low-quality AI reports burdening maintainers.

Big Walk developer House House confirms limited AI use after IGF juror backlash

A group of Independent Games Festival jurors, including Sam Barlow and Lena Raine, issued a statement saying they won't vote for games using generative AI in any capacity, citing a rise in AI-assisted IGF entries. House House, maker of Big Walk, responded by acknowledging it used LLMs sparingly late in development to debug technical issues, estimating under 0.1% of the game's code involved copied LLM snippets.

Proton defends AI-driven coding push after leaked Slack messages spark backlash

Proton published a blog post defending its increased use of AI after a 404 Media report, based on leaked Slack messages, revealed CEO Andy Yen pushing engineers to prioritize speed over traditional security practices like penetration testing. Yen reportedly told staff to 'stop lovingly handcrafting' code and use AI more, dismissing concerns about rushing new code as 'loser thinking.' Proton says its security safeguards remain intact despite the controversy.

Ukraine's fiber-optic drone makers face supply squeeze amid AI data center boom

Ukrainian drone manufacturers rely on Chinese-made fiber-optic cable to build unjammable FPV drones, but rising global demand for optical fiber from AI data centers is driving up costs and tightening supply. Russia and Ukraine together are estimated to produce around 50,000 fiber-optic drones monthly, consuming the cable as a single-use battlefield material, and Chinese customs data reportedly show Russia importing significantly more fiber-optic cable than Ukraine.

Anthropic reportedly targeting $1.3 trillion valuation for IPO

Anthropic, maker of the Claude chatbot, is said to be seeking a valuation of $1.3 trillion as it prepares a stock-market listing. The figure would place the company among the most valuable firms to go public in recent years.

Google DeepMind uses AI agents to formally verify Signal protocol's code

Google DeepMind researchers say they used a new agentic framework called AlphaProof Code to mathematically prove key security properties of the actual code behind the Signal protocol, which underpins encrypted messaging in apps like Signal and Google Messages. The system combined human experts, who defined formal specifications and reviewed select results, with AI agents powered by Gemini Carbon that generated over six million lines of proof code in the Lean language. DeepMind states this is the first time a widely deployed cryptographic protocol's implementation has been formally verified at this depth, by AI or any other method.