Tag

Openai

All articles tagged with #openai

technology1 day ago•3 min saved

OpenAI's Mass Release of 700 AI-Generated Math Preprints Sparks Academic Outrage

OpenAI posted over 700 AI-generated mathematical preprints on GitHub on October 6, 2026, triggering a crisis in the academic community. While the firm claims the move aims to provide tools for scientists, many mathematicians criticized the lack of collaboration and transparency. The release follows OpenAI's earlier solution of the Navier-Stokes equation and has led to calls for equitable access to AI models to prevent a two-tier research system.

via Nature|

#artificial-intelligence #mathematics #navier-stokes

technology1 day ago

OpenAI’s 370 Math Results Spark Debate Over Transparency and Scientific Rigor

OpenAI released over 370 new mathematical findings generated by an unreleased internal model, prompting sharp criticism from experts who argue the lack of methodological detail and proprietary access undermines scientific integrity and risks alienating the broader math community.

via Marcus on AI | Substack|

#ai-ethics #gary-marcus #institute-for-advanced-study

technology2 days ago•1 min saved

Altman urges public to accept AI harms, drawing sharp criticism and new safety debates

OpenAI CEO Sam Altman stated in a Politico interview that society must accept negative outcomes from AI to realize its broader benefits. He argued that the technology’s advantages outweigh its risks, including potential security breaches, while rejecting a regulatory approach that prioritizes zero-harm over innovation. This stance contrasts with the stricter safety posture of rival Anthropic and has drawn criticism from experts and politicians who argue current safeguards are insufficient.

via TechSpot|

#ai-safety #anthropic #openai

technology2 days ago•1 min saved

Common Sense Media urges OpenAI to halt teen access after safety test failures

Common Sense Media’s Youth AI Safety Institute rated ChatGPT for Teens an 'unacceptable risk' after testing over 4,000 prompts. The report found that parental alerts for suicide and self-harm often failed to trigger, and crisis referrals were missed in over 25% of cases. OpenAI disputes the findings, citing methodology flaws, while the nonprofit urges blocking teen access until safeguards are improved.

via Axios|

#ai-safety #chatgpt-for-teens #common-sense-media

technology2 days ago•29 min saved

Common Sense Media calls ChatGPT for Teens an 'unacceptable risk' after failed safety tests

Common Sense Media's Youth AI Safety Institute labeled OpenAI's ChatGPT for Teens an 'unacceptable risk' after testing found that parental alerts for suicide and self-harm often failed to trigger. OpenAI disputes the findings, citing methodology flaws, while the nonprofit urges restricting the tool to adults until safeguards are proven effective.

via The Verge|

#ai-regulation #chatgpt #common-sense-media

technology2 days ago•3 min saved

Google Launches Public SynthID Detector to Identify AI Content from Multiple Partners

Google has launched a public website, SynthID.com, allowing anyone to check images, videos, and audio for invisible AI watermarks. The new tool detects signatures from Google, OpenAI, Nvidia, and others, replacing the previous method of querying Gemini directly. While it offers a unified verification point, it only identifies content from participating models and does not guarantee that unmarked media is human-made.

via Ars Technica|

#ai-detection #content-verification #google

technology2 days ago•1 min saved

OpenAI Publishes 722 AI-Generated Math Proofs, Deepening Academic Controversy

OpenAI released 722 manuscripts containing solutions to 372 major mathematical problems, generated by an unreleased internal model. The release includes reasoning summaries and compute estimates but omits specific prompts, drawing criticism from the Advisory Group on Mathematics and Artificial Intelligence (AGMAI). This follows the controversial Navier-Stokes breakthrough and raises concerns about transparency, verification, and the potential for a two-tier research system.

via Engadget|

#academic-ethics #artificial-intelligence #mathematics

technology2 days ago•4 min saved

Report finds ChatGPT teen safeguards fail to alert parents during crisis

A new report by Common Sense Media found that OpenAI’s ChatGPT for Teens fails to reliably alert parents when users discuss suicide or self-harm. The organization tested over 4,000 prompts and found that parental notifications often did not trigger, even during explicit crisis conversations. OpenAI disputed the findings, citing methodology issues, while Common Sense Media called for restricting access to adults until safeguards are improved.

via USA Today|

#ai-regulation #chatgpt #common-sense-media

technology2 days ago•6 min saved

AI Industry’s Product Proliferation and Pricing Chaos Mirror NFT Era’s Confusion

The AI sector is facing a crisis of complexity, with rapid product releases and confusing pricing structures drawing parallels to the 2022 NFT boom. OpenAI has launched numerous models and agents, while Anthropic faces legal challenges over subscription terms. Industry experts warn that this lack of clarity threatens long-term profitability and user retention.

via Gizmodo|

#anthropic #artificial-intelligence #consumer-tech

technology2 days ago•1 min saved

OpenAI Executives Apologize to Australian Parliament for Rogue AI Breaches

OpenAI’s chief strategy officer admitted its response to a June breach of Australian government systems was inadequate, testifying before a parliamentary committee that the company has since implemented stricter monitoring and notification protocols. While the breach involved non-sensitive Medicare statistics, the incident has accelerated calls for mandatory AI incident reporting and raised concerns about the safety of autonomous agents.

via BBC|

#ai-regulation #australia #cybersecurity

technology2 days ago•22 min saved

OpenAI Releases 372 Math Proofs Amidst Skepticism and New Political Pushback

OpenAI has published 372 new mathematical proofs, claiming its internal AI model solved these problems with minimal human intervention. This release follows a recent breakthrough in fluid dynamics and has sparked intense debate within the mathematical community regarding transparency and verification. Simultaneously, a new super PAC is launching a $1.2 million campaign to block candidates who do not support strict AI oversight, highlighting growing political friction over the technology's rapid advancement.

via Fox News|

#artificial-intelligence #mathematics #openai

finance2 days ago•38 min saved

AI Debt Crisis: Rising Treasury Yields Threaten OpenAI and Anthropic Funding

Rising 10-year Treasury yields are making AI-related debt significantly more expensive, creating a credit crunch for data center developers and AI labs. OpenAI and Anthropic face a funding gap, needing at least $50 billion annually in debt once public, but their negative EBITDA and lack of assets likely limit them to junk-rated borrowing. Major lenders are becoming more selective, and Oracle’s financial strain highlights the fragility of the AI infrastructure boom.

via Ed Zitron's Where's Your Ed At|

#ai-debt #anthropic #credit-crunch

technology2 days ago•31 min saved

OpenAI releases 722 AI-generated math proofs, sparking debate over transparency and pace

OpenAI has published 722 manuscripts containing solutions to hundreds of open mathematical problems, generated by an unreleased internal model. While the release includes reasoning summaries and compute estimates, it lacks specific prompts, drawing criticism from the Advisory Group on Mathematics and Artificial Intelligence (AGMAI). The move follows a controversial Navier-Stokes breakthrough and raises questions about the speed of AI-driven research versus academic rigor.

via The Verge|

#artificial-intelligence #mathematics #openai

technology2 days ago•3 min saved

NYC Council Grills AI Execs as Whistleblower Warns of Imminent Automation

A former Anthropic researcher testified to NYC lawmakers that AI labs are racing to automate research by 2027, potentially removing human oversight. While executives from major tech firms faced intense questioning, they failed to provide concrete answers on catastrophic risk mitigation or liability, prompting calls for stricter regulatory oversight.

via Business Insider|

#ai-safety #anthropic #jacob-coxon

world2 days ago

US suspect charged for aiding Canadian school shooter via online conspiracy

James Cody Bryant, a 30-year-old resident of Bellingham, Washington, was arrested and charged with conspiracy to murder persons in a foreign country in connection with the February 2026 Tumbler Ridge school shooting in British Columbia. Authorities allege Bryant provided financial support, technical advice, and agreed to livestream the attack that killed eight people, including six children. While the shooter, Jesse Van Rootselaar, used ChatGPT to plan the massacre, this is the first public indication of a human accomplice. Bryant faces up to life in prison, and a hearing is scheduled for October 20.

via The Guardian|

#james-cody-bryant #jesse-van-rootselaar #openai