General Cyber News via Ars Technica Risk Assessment

OpenAI's new GPT‑5.3‑Codex‑Spark is 15 times faster at coding than its predecessor.
Publish Date: 2/12/2026
read more -->

Distillation technique lets copycats mimic Gemini at a fraction of the development cost.
Publish Date: 2/12/2026
read more -->

ClickFix bait, combined with advanced Castleloader malware, is installing Lumma "at scale."
Publish Date: 2/11/2026
read more -->

Zoë Hitzig resigned on the same day OpenAI began testing ads in its chatbot.
Publish Date: 2/11/2026
read more -->

The $20,000 experiment compiled a Linux kernel but needed deep human management.
Publish Date: 2/6/2026
read more -->

Incident is at least the third time the exchange has been targeted by thieves.
Publish Date: 2/6/2026
read more -->

Claude Opus 4.6 and OpenAI Frontier pitch a future of supervising AI agents.
Publish Date: 2/5/2026
read more -->

Sam Altman calls AI competitor "dishonest" and "authoritarian" in lengthy post on X.
Publish Date: 2/5/2026
read more -->

Publishers are rolling out more aggressive defenses.
Publish Date: 2/5/2026
read more -->

The window to patch vulnerabilities is shrinking rapidly.
Publish Date: 2/4/2026
read more -->

ChatGPT competitor comes out swinging with Super Bowl ad mocking AI product pitches.
Publish Date: 2/4/2026
read more -->

Some semi-unhinged musings on where LLMs fit into my life—and how I'll keep using them.
Publish Date: 2/4/2026
read more -->

Two AI giants shake market confidence after investment fails to materialize.
Publish Date: 2/3/2026
read more -->

We don't need self-replicating AI models to have problems, just self-replicating prompts.
Publish Date: 2/3/2026
read more -->

Suspected China-state hackers used update infrastructure to deliver backdoored version.
Publish Date: 2/2/2026
read more -->

Moltbook lets 32,000 AI bots trade jokes, tips, and complaints about humans.
Publish Date: 1/30/2026
read more -->

Ars spoke to several software devs about AI and found enthusiasm tempered by unease.
Publish Date: 1/30/2026
read more -->

Settlement comes more than 6 years after Gary DeMercurio and Justin Wynn's ordeal began.
Publish Date: 1/29/2026
read more -->

We have no proof that AI models suffer, but Anthropic acts like they might for training purposes.
Publish Date: 1/29/2026
read more -->

One of the last holdouts for ransomware discussions, RAMP is taken down.
Publish Date: 1/28/2026
read more -->