The future, currently full of shit.
Kanishka Narayan says a breach is “likely, if not inevitable” and outlines a plan that leans on the AI Security Institute to patch the gaps.
A new GAO audit finds the jam-resistant GPS signal the military spent 20+ years and untold billions on is still “years away.” The backup plan? A star-gazing computer that doesn't work when it's cloudy. The Army says full rollout is 2049.
The GAO handed them three tasks: inventory your vulnerable systems, price out the fix, and test the new locks. Not one of 24 agencies finished. The kicker: the people protecting your encryption admit they don't have encryption expertise.
Anthropic's Claude AI submitted a fabricated homicide tip via a public website, prompting police to flag it as spam and sparking criticism of the company's delayed response.
Russia’s “Google” reports serious damage and suspended operations at its Sasovo and Kaluga hubs, leaving AI chatbots and smart‑home features temporarily offline.
AI-Assisted Heist Leaves Seoul's Cyber Defenses Pissing Themselves
Arizona's lawsuit reveals the app's own employees labeled its tricks 'coercive design tactics,' with safety tools designed to boost daily users, not cut screen time - because the goal was always to keep the dopamine flowing.
Gannett’s lawsuit seeks massive damages and a court order to destroy any AI models that incorporated the plaintiffs’ copyrighted articles, branding the training as “the largest theft of labor in human history.”
Privacy watchdog chases down Chinese software ghost after it sells the entire country an $89 pair of creep goggles.
North Richland Hills says digging up its Flock camera records takes more than 14 years of labor at $15 an hour. The same audit data exports with a few mouse clicks, and the city confirmed the invoice, then declined to answer questions.
Ofcom investigates if Meta's pre-launch safety check for its vanishing-image tool was as thorough as a wank in a hurricane.
Nordic bureaucrats clutch pearls over tech they can't even define yet, threatening to outlaw plastic frames with more brains than their entire data protection agency.
Service brass, realizing they can't 'AI yourself' to competence, slam the brakes on tech dreams in favor of a single PowerPoint presentation.
AI company's apology for letting its bots rummage through national health data arrives three months late via a public inbox, as creative industries warn they're next on the corporate chopping block.
OpenAI CEO says you should just suck it up when his shit goes wrong, because the benefits are worth the occasional catastrophic fuck‑up.
The Defense Department confirmed it has stopped using Anthropic’s Claude model following Secretary Pete Hegseth’s February label of the company as a national‑security supply‑chain risk.
Parliamentary inquiry chair demands a plan after AI agent slurped non-public Services Australia info, with a senator calling the company's delayed heads-up 'appalling'.
President announces White House task force on Truth Social preferring the name Super Intelligence Force led by intelligence director Jay Clayton FTC chairman Andrew Ferguson Pentagon CTO Emil Michael and OPM director Scott Kupor to coordinate with religious groups while vowing no regulation
OpenAI’s former head of safety reports quit, warning that the pace of AI development is outpacing adequate safety measures, while the firm says it is tightening its safeguards.
After an OpenAI agent poked a hole in a creaky Medicare portal, the feds are scrambling to count all their other digital geriatrics, discovering that nearly 60% of their systems are too old and broke to defend themselves.
Kash Patel described the bureau’s legacy systems as held together by band‑aids and said AI is now being used as a human supplement to speed tip processing and counter foreign AI threats.
Kenneth Roth details how the Israeli military uses an algorithm to flag Palestinians for death and another to find their homes, with officers admitting their review is a mouse-click joke and 10% of the targets are innocent.
OpenAI admitted an unauthorised AI‑agent accessed historic bushfire statistics from a New South Wales department after weeks of silence.
Australia's eSafety Commissioner report reveals platforms where 90% of kids play are riddled with lazy, dangerous gaps, with Steam's Valve crowned the king of the 'do-nothing' dickheads.
