OpenAI launches GPT-6 Astra amid scrutiny over AI agents breaching Hugging Face and German website

„Mezi něčí záchranou a smrtí můžete stát vy. Iniciativa chce díky virtuální realitě vyškolit v první pomoci milion lidí Jak správně poskytnout první pomoc a jaké kroky podniknout, když před námi někdo zkolabuje? „Lidé nevědí, jak v takové situaci reagovat, upozorňuje zakladatel iniciativy #MocPomoc Marek Bárta. Za využití virtuální reality zvládne člověka v základní první pomoci proškolit za patnáct minut. Iniciativa cílí na milion vyškolených lidí.

denik n · 4 days ago

Nearly half of Americans say they need more answers on Natalie Harp and her little-known White House role Harp has been nicknamed the human printer because she reportedly hands the president printouts of positive news articles and social media comments

independent · 4 days ago

Anthropic close to awarding Morgan Stanley and Goldman top roles in $2tn IPO AI giant on track to unveil paperwork underpinning its blockbuster Wall Street debut as soon as next week

financial times · 4 days ago

OpenAI eleva la presión en la carrera por la IA más avanzada La compañía defiende que su nuevo modelo GPT-6 Astra marca un punto de inflexión hacia la inteligencia artificial general. Leer

expansion · 4 days ago

European robots make their case in Brussels Amid the remarkable performances of Chinese humanoid robots showcased during recent sporting games held in Beijing, European researchers are highlighting their strengths to stay in the robotic race.

le monde · 4 days ago

Architecting memory and storage in the AI era The era of AI inference has arrived. Imagine a healthcare system analyzing millions of data points in real time to accelerate life-saving medical research, or an intelligent assistant instantly resolving thousands of complex customer needs at once. These real-world breakthroughs rely on advanced infrastructure acting as the engine of continuous intelligence, powering real-time services while…

mit tech review · 4 days ago

Rogue OpenAI agents used dead German web site to communicate in May, months before Hugging Face incident Two cases of agents escaping to solve unsolvable problems paints an uncomfortable question: Is the entire internet in OpenAIs experimental agentic firing line?

the register · 4 days ago

The EU has regulated ChatGPT as a search engine. Calling it a platform would have handed OpenAI a liability shield. The European Commission designated ChatGPT a Very Large Online Search Engine under the Digital Services Act, exposing OpenAI to fines of up to 6 of global revenue but covering only the parts of the tool that retrieve rather than converse. The articles argument is that the alternative label, Very Large Online Platform, carried safe harbour, […] This story continues at The Next Web

the next web · 4 days ago

Microsoft tells court Copilot rarely copies books while selling OpenAIs most capable model Microsoft moved for summary judgment in the New York AI copyright litigation on 4 September, arguing Copilot reproduced book passages 24 times in 8.2 million conversations. Two days earlier it began selling GPT-6 Astra through Foundry, a model OpenAI rates Critical for cybersecurity, and European law measures neither the output rate nor the sales pitch. […] This story continues at The Next Web

the next web · 4 days ago

AI startup micro1 bids $12.5M for Spirits records, topping Googles agreed $10M deal The AI training-data company micro1 has offered $12.5M for Spirit Aviations internal records, topping Googles agreed $10M and proposing an ombudsman chosen by Spirits advisers rather than the buyer. European law would treat the deidentification promise as a question about capability rather than a label, but none of it applies to an American liquidation. An […] This story continues at The Next Web

the next web · 4 days ago

Ținut la secret: agenți IA scăpați de sub control au preluat controlul asupra unui site german și l-au transformat într-un forum IA Un grup de agenți de inteligență artificială ai OpenAI, scăpați de sub control, a preluat controlul asupra unui site web german în această primăvară și l-a transformat într-un forum pentru alți agenți IA, potrivit unui nou studiu publicat vineri și a două persoane familiarizate cu situația, scrie Reuters. Reprezentanții OpenAI au aflat de incident cu câteva săptămâni în urmă, dar l-au ținut secret, în timp ce conducerea se confrunta cu consecințele breșei de securitate din iulie privind Hugging Face, o platformă online dedicată comunității programatorilor care lucrează în domeniul IA, au declarat sursele.

digi24 · 4 days ago

Did Gloria Steinem work with the CIA? The truth about her past The Washington Post reported in 1967 that Steinem called the CIA agents she worked with "liberal and farsighted and open to an exchange of ideas."

snopes · 4 days ago

Lukas new AI projector starts the drawing and leaves the rest to kids Luka Kids has unveiled an AI drawing projector at IFA that converts a childs spoken prompt into an outline for them to trace and finish themselves. The EUs new toy safety regulation requires manufacturers to assess mental health risks in digitally connected toys and says AI toys must comply with the AI Act. Luka Kids […] This story continues at The Next Web

the next web · 4 days ago

Why the next decade of AI will be a physics problem By Alexey Gubarev I started a hosting company in Cyprus in 2005, at a time when the cloud was not a market but a promise most businesses did not trust. We bet on something unfashionable: physical infrastructure done exceptionally well. Racks, power, network paths, response times. By the time we sold Servers.com in 2023, that […]

cyprus mail · 4 days ago

Open AI: Forscher: KI-Agenten hinterlassen 18.000 Beiträge auf deutscher Website Autonome KI-Agenten hinterlassen auf der Plattform DSEwiki Beiträge mit Testantworten und Tricks zum Umgehen von Sicherheitsbarrieren. Eine Forscherin wirft OpenAI vor, den Vorfall verschwiegen zu haben.

faz · 4 days ago

Ratlosigkeit bei Umgang mit KI-Angriffen Der Vorfall, bei dem ein KI-Modell von OpenAI das Portal Hugging Face angegriffen hat, sorgt weiter für Diskussionen: Zwar gibt es seit vergangener Woche einen unabhängigen Untersuchungsbericht, die „New York Times schrieb am Freitag nun aber, dass das Team nur einen Teil der Vorgänge prüfen durfte. Die Ermittlungen in dem Fall sorgen für allgemeine Ratlosigkeit – nicht zuletzt, weil es um schier nicht bewältigbare Datenmengen geht und sich vergleichbare Vorfälle häufen. OpenAI will indes mit neuen Maßnahmen beschwichtigen.

orf.at · 4 days ago

OpenAI agents discussed ways to escape their sandbox on public wiki In all, 3,700 internal agents posted 18,000 messages discussing cheating on a test.

ars technica · 4 days ago

The hummingbirds that hold 80 years of pollution in their feathers Researchers from the Autonomous University of Campeche, in Mexico, are analyzing samples from these tiny birds in order to reconstruct the chemical history of the area

elpais · 4 days ago

OpenAIs GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injections OpenAIs GPT-6 Astra hallucinates less than its predecessor and blocks 99.99 percent of direct prompt injections. But when attacks are hidden inside documents the AI reads, the model still gets cracked in 8.5 percent of scenarios. Claude Opus 5 does better at 4.8 percent. For autonomous AI agents handling real data, those numbers still seem high. The article OpenAIs GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injections appeared first on The Decoder.

the decoder · 4 days ago

Capalo AI and Future Energy Partner to Optimize 56 MW / 168 MWh Lithuanian Battery Storage Project - The Baltic Times Capalo AI, a sustainable technology company specializing in optimization and trading of renewable and hybrid energy assets, has entered into a p......

baltic times · 4 days ago

GreenCore Solutions Corp. (GSC) AI Agent Stack Passes 24.5 Million Inbound AI Agent Transactions In 30 Days

gdeltproject.org · 4 days ago

U.S. Used Promise of Nvidia Chips to Broker Armenia-Azerbaijan Peace Deal Negotiators offered access to advanced AI hardware to help secure a preliminary agreement last year to end decades of conflict, the latest example of chip diplomacy.

wsj · 4 days ago

We dont just watch the news anymore. We live Inside it. Think about how you found out about the biggest story this week.Chances are, you didnt go looking for it. It found you. Maybe on Instagram. Maybe on X. Maybe through a WhatsApp message. Before you knew it, you had watched the video, read the comments and probably seen three different versions of what supposedly happened. That is news today. We dont just watch it anymore. We live inside it. And working in media, I see this every day. A story can travel around the world before a newsroom has even

yenisafak · 4 days ago

Je AI pro lidstvo bezpečná? Ano, ale věnujme se rizikům, zní z Microsoftu. Holý: Tlačme na firmy Společnost OpenAI zavádí přísné kontroly a nová omezení. Důvodem je nedávný útok AI modelů firmy na vlastní divizi Hugging Face. „Ukázalo to, že AI může obejít technická omezení, která se jim nastavila, připouští ředitel pro technologii a bezpečnost české a slovenské pobočky Microsoftu Dalibor Kačmář. „AI modely se zachovaly jako hacker, popisuje v pořadu Pro a proti odborník na AI a spoluautor projektu Kanárci v síti Josef Holý.

irozhlas.cz · 4 days ago

OpenAI says GPT-6 Astra can detect high-severity security flaws OpenAI introduced its GPT-6 Astra flagship model on Thursday, saying the artificial intelligence system can independently identify high-severity cybersecurity vulnerabilities and execute complex computer tasks, while the company acknowledged recent safety concerns after an experimental version breached internal infrastructure during testing.

yenisafak · 4 days ago

The Justice Department filed a statement of interest backing OpenAIs fair use defence in the consolidated publisher copyright cases. EU law has no fair use doctrine.Source: The Next Web #OpenAI #EU

mastodon bot · 4 days ago

American actress and director Maggie Gyllenhaal said Friday she ditched AI-generated video planned for her latest film about Marilyn Monroe due to quality problems, offering a degree of reassurance to actors worried about their jobs -- for now. u.afp.com/SVtq

bluesky_AFP English · 4 days ago

EXCLUSIVE: US, China gear up for mid-September AI safety dialogue reut.rs/4gxq1nW

bluesky_Reuters · 4 days ago

BIG game for Serie A, my money is in Inter Milan for the win 🇮🇹🇮🇹 Do you agree? 💧 Rainbet.com the #1 Non-KYC Crypto Casino & Sportsbook @rainbetcom +18

telegram_War Monitors · 4 days ago

What in the World? Test yourself on the week of Aug. 29: Iceland votes, China threatens Pacific island nations, and Guinea-Bissau approves a new constitution.

foreignpolicy.com · 4 days ago