GoKawiil
Tech News
OpenAI reportedly used an AI-based analysis tool to scan the Linux kernel and identified 948 possible local privilege escalation vulnerabilities, according to Phoronix founder Michael Larabel. The findings point to weaknesses that could let a local user gain elevated system access if left unaddressed.
In San Francisco, some dating app users are adding statements about artificial intelligence to their profiles, including explicit rejections such as 'Fuck AI' or warnings not to date them if they use AI. Interviews across parties, coworking spaces and dating profiles found people treating AI attitudes as a values signal similar to political affiliation, while others wrestle with the ethics and practicalities of dating someone who works in the AI industry.
On the latest Vergecast episode, hosts Nilay Patel and David Pierce discuss how Democratic gains in the 2026 midterms could reignite congressional attention on tech CEOs, AI regulation, and data center buildouts. They also examine whether AI companies, including OpenAI and Google, have found sustainable business models despite AI search growth and reported pullbacks in enterprise spending.
Researcher Derik Kauffman discovered a hidden list of over 100,000 words and phrases blocked from iOS autocomplete and swipe-typing across major languages, and published it online. The list includes slurs, sexual and drug terms, but also political phrases like 'dictator Trump,' 'sleepy Joe,' and 'covid was a hoax,' plus inconsistencies such as blocking 'bad girl' but not 'bad boy.' Kauffman extracted the list from a hash table buried in iOS using AI coding tools, and another researcher independently verified the findings; Apple told Reuters the keyboard avoids suggesting potentially offensive terms but doesn't stop users from typing them manually.
OpenAI uploaded more than 700 AI-generated mathematical manuscripts to a public preprint repository in early October, claiming new mathematical results pending verification. Mathematicians have criticized the papers as sloppy and hard to verify, noting no human author is credited to take responsibility for the arguments, and OpenAI has not signed the Leiden Declaration on Artificial Intelligence and Mathematics, which calls for clear citation and disclosure practices.
OpenAI has published a National Community Compact and Compute Campus Policy Blueprint outlining commitments for every future compute campus it builds. The company pledges to pay full energy and infrastructure costs, provide financial assurances to other utility customers, publish water and noise impact data, avoid NDAs with public officials, and report annually on local hiring and investment.
OpenAI announced it will stop using non-disclosure agreements with public officials on its data center projects, limiting confidentiality to narrow cases like security or trade secrets. The company also committed to paying the full cost of electricity and infrastructure it uses, publishing job creation estimates, and attaching a 'local compact' of commitments to every future data center, which it calls a 'compute campus.'
Apple's annual financial filing added a risk factor warning that AI agents capable of acting autonomously could behave in 'unintended or unpredictable ways,' potentially exposing the company to lawsuits or reputational harm. The filing notes such agents can interact with systems without authorization, alter code, send communications or complete transactions without explicit direction.
OpenAI has brought on security personnel with prior experience at the Department of Government Efficiency (DOGE), including individuals previously involved in breaching Anthropic's systems, to bolster its security operations. The company also announced it is raising rewards for critical vulnerability findings reported through its bug bounty program to more than four times their previous level.
OpenAI announced a Grid Protectors Program designed to cut electricity use at its data centers when local grids are under heavy demand, rather than waiting for new power supply to come online. The company says it has already signed a 3.2-gigawatt agreement with Georgia Power for its planned Effingham County campus, including 1,000 megawatts set aside for flexible demand response.
A developer reported that OpenAI's Codex tool had its 'Improve the model for everyone' data-sharing setting turned back on without their knowledge, despite having previously disabled it. Coworkers confirmed the same issue, and the poster noted the setting can be checked and disabled via OpenAI's data-controls and privacy pages. Other users on Hacker News had flagged similar behavior before.
According to Axios, OpenAI is preparing a policy and public-relations campaign that will refer to data centers as 'AI compute campuses,' launching after the US midterm elections. OpenAI chief global affairs officer Chris Lehane told Axios the company wants to shape public perception through messaging, emphasizing claims that such facilities bring jobs and benefits to host communities.