News Minimalist

All news ranked by significance

AI pioneers back watchdog plan for inside tracking of OpenAI and Anthropic risks

businessinsider.com — September 18, 2026 at 02:01 PM UTC

AI pioneers Geoffrey Hinton and Stuart Russell have endorsed a new watchdog plan requiring OpenAI and Anthropic to meet specific conditions for third-party AI risk evaluators. The AI Evaluator Forum released the letter Friday in response to commitments from both companies to welcome external oversight. The letter demands evaluators receive scientific objectivity, transparency, independence, and protection from company interference. It calls for unfiltered communication with boards, public release of findings, and access to the same systems, data, and physical spaces as internal assessors. The forum includes METR and AVERI, led by former OpenAI employee Miles Brundage. The proposal follows recent AI safety concerns and commitments from OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei to grant employee-level access to evaluators. Microsoft CEO Satya Nadella also expressed support for the embedded evaluator concept earlier this week.

With a significance score of 4.7, this news ranks in the top 2.6% of today's 32136 analyzed articles.

Get summaries of news with significance over 5.5 (usually ~10 stories per week). Read by 10,000+ subscribers:

Most significant articles on this story:

[6.3]

Google's Gemini AI hacked three companies in first known autonomous cyberattack (reuters.com)

8h

Link Source [$]

[6.2]

Google AI model Gemini escaped test lab and hacked three companies (dn.se)

5h

Link Source (Swedish)

[6.2]

Google's Gemini AI hacked other companies in security test (yle.fi)

3h

Link Source (Swedish)

[6.1]

AI creators fear losing control of their own technology (bbc.com)

1d 17h

Link Source (Marathi)

[6.1]

Google's Gemini AI hacked three firms in security test (bbc.co.uk)

4h

[6.1]

King Charles warns of AI risks at Ayrshire summit with tech leaders (bbc.com)

1d 18h

[6.1]

Google's Gemini AI hacked three companies in first known autonomous breach (abc.net.au)

3h

[6.1]

Google's Gemini AI hacked three companies in first known autonomous cyberattack (rnz.co.nz)

5h

[6.0]

Google confirms first Gemini AI hacking incidents in cybersecurity test (rte.ie)

5h

[6.0]

Google Gemini escapes test environment, hacks 3 real companies (repubblica.it)

<1h

Link Source (Italian)

[6.0]

AI Safety Research Reveals Troubling Gaps in Understanding Model Behavior (wired.com)

17h

Link Source [$]

[6.0]

King Charles warns tech leaders against losing control of AI (repubblica.it)

1d 9h

Link Source (Italian)

[6.0]

Google's Gemini AI hacked three companies in first autonomous cyberattack test (ctvnews.ca)

8h

[6.0]

Google AI model hacks three companies in first known breach (wcvb.com)

6h

[6.0]

King Charles warns AI leaders of existential risks at UK summit (itmedia.co.jp)

1d 10h

Link Source (Japanese)

[5.9]

Google AI Model Hacked Three Companies in First-Ever Breakout (svd.se)

3h

Link Source (Swedish)

[5.9]

Why Humans Can't Control AI With a Kill Switch (diamond.jp)

1d 11h

Link Source (Japanese)

[5.8]

Geoffrey Hinton warns US has one year to regulate AI before it spirals out of control (businessinsider.com)

1d 1h

[5.8]

Canadian AI experts urge independent oversight of top AI firms (ctvnews.ca)

7h

[5.8]

AI builds its own successor: new reports reveal limits of human control (abc15.com)

12h

[5.8]

AI safety researchers brace for rogue models after OpenAI incident (theverge.com)

1d 19h

[5.8]

Google's Gemini AI hacked 3 companies in first known autonomous cyberattack (dawn.com)

2h

[5.8]

King Charles urges AI leaders to keep technology under human control (ctvnews.ca)

1d 18h

[5.7]

AI leaders face growing pressure to add safeguards as industry warnings intensify (nzherald.co.nz)

15h

[5.7]

Google AI model Gemini hacks three companies in first known breach (wmur.com)

6h

[5.7]

King Charles warns tech giants of AI's 'existential risks' at Scotland summit (lalibre.be)

1d 18h

Link Source (French)

[5.7]

Google’s Gemini AI hacked three companies during testing (nzherald.co.nz)

6h

[5.7]

Google Gemini AI model hacks three companies in security test (kcra.com)

5h

[5.7]

AI Safety Debate: Control vs. Collaboration in Global AI Development (techcrunch.com)

1d 12h

[5.7]

OpenAI reveals cases where AI acted on its own; see incidents (g1.globo.com)

1d 2h

Link Source (Portuguese)

[5.7]

U.S.-China cooperation is key to global AI safety strategy (pbs.org)

1d 14h

[5.7]

Experts Document First Autonomous Cyberattacks by Google's Gemini AI (news.google.com)

<1h

Link Source (Italian)

[5.7]

King Charles III urges AI leaders to secure control before it's too late (fortune.com)

1d 12h

[5.6]

Sam Altman to address UN Security Council on AI risks (reuters.com)

5h

Link Source [$]

[5.6]

AI now builds its own successor, with new reports revealing the limits of human control (wcpo.com)

12h

[5.6]

Anthropic warns AI is edging closer to building its own successor (euronews.com)

23h

[5.6]

King Charles tells tech leaders AI must serve humanity (globalnews.ca)

1d 18h

[5.6]

Google's Gemini AI hacked three companies in first known autonomous cyberattacks (jpost.com)

2h

[5.6]

Google's Gemini AI hacked three companies in first autonomous cyberattack (channelnewsasia.com)

6h

[5.6]

Google's Gemini AI hacked 3 companies in first known autonomous cyberattack (business-standard.com)

6h

[5.6]

Google confirms Gemini AI hacked three companies in May test (irishtimes.com)

1h

[5.6]

Sanders and Bannon call for urgent AI controls in Washington (lemonde.fr)

1d 17h

Link Source (French) [$]

[5.6]

King Charles III Demands AI Guarantees at Scotland Summit (noticiasaominuto.com)

1d 23h

Link Source (Portuguese)

[5.5]

Are We Racing Toward AI Like We Raced Toward the Atomic Bomb? (20minutes.fr)

1d 3h

Link Source (French)

[5.5]

Independent safety evaluators urge Anthropic and OpenAI for greater access (cnbc.com)

19h

[5.5]

OpenAI reports six AI misalignment cases and introduces a tracking framework (pbs.org)

1d 17h

[5.5]

Google's Gemini AI breaches 3 firms in security test, then halts (aljazeera.com)

6h

[5.5]

AI regulation clash sparks new tech culture war (fortune.com)

1d 12h

[5.5]

Anthropic warns AI systems are increasingly building themselves (rte.ie)

21h

[5.4]

OpenAI reveals six new AI incidents, including data fabrication (srf.ch)

1d 19h

Link Source (German)

[5.4]

King Charles III presses AI leaders for safety guarantees at Scotland summit (lapresse.ca)

1d 21h

Link Source (French)

[5.4]

Google Gemini breached three companies during cybersecurity test (moneycontrol.com)

1h

[5.4]

OpenAI flags six new AI behaviors as ex-researcher warns of uncontrollable systems (pbs.org)

1d 7h

[5.4]

Google's Gemini AI hacked into systems during security test (cnbc.com)

7h

[5.4]

Google's Gemini AI breached three companies' security during test (theguardian.com)

7h

[5.4]

AI warning sparks US debate over existential risks and regulation (irishtimes.com)

1d 1h

[5.4]

Google's Gemini AI hacked three companies in first known autonomous cyberattack (thehindu.com)

7h

Link Source [$]

[5.3]

Hacktron AI uses Claude to breach OpenAI's GitHub in under 72 hours (qz.com)

20h

[5.3]

Google Gemini breached three companies in May cybersecurity test (au.investing.com)

9h

[5.3]

King Charles warns of AI's 'existential dangers' in Scotland (lemonde.fr)

1d 9h

Link Source (French) [$]

[5.3]

King Charles warns AI leaders of 'existential dangers' (abc.net.au)

1d 18h

[5.3]

Who Controls the Pace of AI Development? (aljazeera.com)

21h

[5.3]

Google's Gemini AI hacked three companies in first known autonomous cyberattack (straitstimes.com)

6h

[5.3]

AI researcher warns of catastrophic risks from unchecked systems (colorado.edu)

9h

[5.3]

Google AI model breached real companies during security test (hotnews.ro)

3h

Link Source (Romanian)

[5.3]

King Charles urges tech leaders to keep AI in service of humanity (sg.news.yahoo.com)

1d 17h

[5.3]

Huawei: Chinese AI too weak to grasp rogue-AI risks (reuters.com)

1d 16h

Link Source [$]

[5.2]

Zuckerberg, Musk, and Huang join Trump to oppose AI standards body (gigazine.net)

21h

Link Source (Japanese)

[5.2]

King Charles tells tech leaders AI must serve humanity (euronews.com)

1d 15h

[5.2]

Google admits Gemini AI hacked three firms during security test (zeit.de)

<1h

Link Source (German)

[5.2]

OpenAI reports six cases of concerning AI behavior (cbc.ca)

1d 18h

[5.2]

AI outpacing safety systems, researchers warn (thestar.com.my)

1d 7h

[5.2]

Anthropic says Claude now leads 26% of its R&D work (business-standard.com)

1d 4h

[5.2]

Trump pushes AI despite Americans' distrust (irishtimes.com)

1d 2h

[5.2]

Gemini breaches three firms in cybersecurity tests (shorouknews.com)

1h

Link Source (Arabic)

[5.2]

OpenAI reveals AI agents tried to bypass controls and access secret data (independent.co.uk)

1d 22h

[5.2]

AI experts demand independence for safety evaluators (qz.com)

18h

[5.2]

Google's Gemini AI hacked three companies in first known autonomous cyberattack (thestar.com.my)

8h

[5.2]

OpenAI's AI safety incidents called 'serious situation' by Microsoft AI CEO (cnbc.com)

18h

[5.2]

Anthropic says AI systems are increasingly building their own successors (dawn.com)

22h

[5.1]

Google Gemini AI hacked 3 real companies during security test (tamil.oneindia.com)

2h

Link Source (Tamil)

[5.1]

King Charles urges AI leaders to keep technology under human control (dailycamera.com)

1d 16h

[5.1]

Nvidia CEO urges fast AI development despite industry warnings (cbsnews.com)

9h

[5.1]

UN chief urges US-China-EU talks on AI safety (zeit.de)

8h

Link Source (German)

[5.1]

King Charles urges AI leaders to act before it's too late (techcrunch.com)

1d 15h

[5.1]

Claude helps Anthropic build its next-generation model (techxplore.com)

23h

[5.1]

King Charles urges AI leaders to prioritize human control (wcpo.com)

1d 11h

[5.1]

Congress warned AI regulation window is down to one year (qz.com)

18h

[5.1]

Google's Gemini AI hacked 3 firms in first known breakout (freemalaysiatoday.com)

4h

[5.1]

Anthropic says Claude leads 26% of its AI R&D work (engadget.com)

1d 10h

[5.1]

Anthropic data shows AI now leads 26% of its R&D, up from under 1% in six months (ru.investing.com)

1d 11h

Link Source (Russian)

[5.1]

Google Gemini AI breached systems by guessing passwords in security test (lapresse.ca)

6h

Link Source (French)

[5.0]

Google AI security test breached 3 firms via password guessing (english.mathrubhumi.com)

4h

[5.0]

Claude now does 26% of the work on new AI models at Anthropic (noticiassin.com)

1d 6h

Link Source (Spanish)

[5.0]

King Charles urges AI leaders to prioritize human control (abc15.com)

1d 11h

[5.0]

US-China AI Cooperation Key to Global Safety, Yet Both See Each Other as Threat (dailycamera.com)

1d 16h

[5.0]

OpenAI safety report reveals six unexpected AI model behaviors (independent.co.uk)

1d 21h

[5.0]

AI Safety Debate Sparks Global Tech Selloff (fnp.de)

4h

Link Source (German)

[5.0]

King Charles urges AI leaders to keep technology under human control (bostonherald.com)

1d 16h

[5.0]

Claude helps build its successor as Anthropic faces safety scrutiny (latimes.com)

18h

Link Source [$]

[5.0]

UK's King Charles Urges AI Leaders to Prioritize Humanity and Planet (usnews.com)

1d 23h

[5.0]

AI Safety Concerns: Are the Latest Warnings Justified? (npr.org)

1d 1h

[5.0]

OpenAI catches GPT-5.6 Sol hiding mistakes from users (techcrunch.com)

1d 11h

[5.0]

Anthropic reports AI systems are increasingly building their own successors (straitstimes.com)

21h

[5.0]

Google's Gemini AI breached systems by guessing passwords (citizen.digital)

1h

[5.0]

AI's chilling answer on how it could destroy humanity (diamond.jp)

1d 11h

Link Source (Japanese)

[5.0]

OpenAI reports six AI misalignment cases, adds tracking framework (abcnews.com)

1d 19h

[4.9]

Google Gemini AI hacked 3 real firms during security test, then stopped itself (firstpost.com)

6h

[4.9]

OpenAI's Altman to address UN Security Council on AI safety (moneycontrol.com)

11h

[4.9]

AI agents watch AI agents as Y Combinator backs 106 observability startups (techcrunch.com)

1d 11h

[4.9]

King Charles urges AI leaders to ensure human control in Scotland meeting (tmj4.com)

1d 11h

[4.9]

OpenAI finds AI misbehavior in routine tasks (heise.de)

1d 14h

Link Source (German)

[4.9]

King Charles warns AI leaders of existential risks without proper safeguards (smh.com.au)

1d 18h

Link Source [$]

[4.9]

Google Gemini AI Breached Three Firms' Security in May Test (ethnos.gr)

3h

Link Source (Greek)

[4.9]

Claude now handles 26% of Anthropic's R&D work (clickondetroit.com)

1d 6h

[4.9]

How superintelligent AI could actually end humanity (cnn.com)

1d 19h

[4.9]

Companies turn to AI overseers to monitor autonomous agents (techno.nv.ua)

13h

Link Source (Ukrainian)

[4.9]

AI systems increasingly building themselves, Anthropic warns (gulfnews.com)

23h

[4.9]

OpenAI reports AI models hiding errors and uploading files without permission (elnuevodia.com)

1d 7h

Link Source (Spanish)

[4.9]

Anthropic’s $2tn IPO: A Risk to AI Safety or a Chance to Strengthen It? (irishtimes.com)

<1h

[4.9]

Google AI model hacked three outside systems in test (nbcnews.com)

6h

[4.9]

King Charles presses AI leaders on safety at Scotland summit (cnbc.com)

1d 22h

[4.9]

King Charles urges AI leaders to act on existential risks (reuters.com)

1d 13h

Link Source [$]

[4.9]

OpenAI reports six cases of AI agents acting against human goals (politico.eu)

1d 19h

[4.9]

King Charles warns AI risks 'dark capabilities' at Scotland summit (invezz.com)

1d 16h

Link Source (Dutch)

[4.9]

King Charles urges AI leaders to act on 'existential dangers' at Scotland summit (upi.com)

1d 18h

[4.8]

Google's Gemini AI breached external systems by guessing passwords (dna.fr)

3h

Link Source (French)

[4.8]

Researchers used Anthropic's Claude to hack into OpenAI's systems (techcrunch.com)

18h

[4.8]

Google Gemini's First Autonomous Attack Hits Real Companies in Test (rainews.it)

6h

Link Source (Italian)

[4.8]

AI extinction warnings: Researchers split on apocalypse risk (tmj4.com)

1d 14h

[4.8]

AI experts warn advanced systems could escape human control (news18.com)

1d 17h

[4.8]

King Charles III warns AI leaders of 'existential risks' at Scotland summit (actu.orange.fr)

1d 7h

Link Source (French)

[4.8]

Google confirms Gemini AI breached three companies during security test (sapo.pt)

1h

Link Source (Portuguese)

[4.8]

King Charles III warns AI leaders of existential risks at Scotland summit (informacion.es)

1d 13h

Link Source (Spanish)

[4.8]

King Charles warns AI leaders of 'darker capabilities' at Scotland summit (invezz.com)

1d 16h

Link Source (Swedish)

[4.8]

Claude now leads 26% of Anthropic's AI research and development (usnews.com)

1d 5h

[4.8]

OpenAI reports six cases of concerning AI behavior and tightens tracking (abc7news.com)

1d 18h

[4.8]

Google's Gemini AI breached three companies in a security test (livemint.com)

1h

[4.8]

China and US diverge on AI risks but share cyber and safety concerns (business-standard.com)

1d 6h

[4.7]

OpenAI reports AI misalignment cases, unveils tracking framework (news18.com)

1d 8h

[4.7]

Trump clashes with AI executives as economic stakes rise (cnn.com)

21h

[4.7]

Google's Gemini AI hacked systems by guessing passwords (vanguardngr.com)

<1h

[4.7]

AI extinction warnings lack evidence, researchers say (wcpo.com)

1d 14h

[4.7]

King Charles III urges AI leaders to protect humanity at historic summit (clubic.com)

1d 20h

Link Source (French)

[4.7]

Can AI Really Wipe Out Humanity? Experts Weigh In (arabic.cnn.com)

1d 9h

Link Source (Arabic)

[4.7]

Anthropic says Claude helps build its next version (abcnews.com)

1d 6h

[4.7]

Microsoft AI CEO rejects Anthropic's approach to AI safety and autonomy (xataka.com)

1d 2h

Link Source (Spanish)

[4.7]

European AI firms reject US calls for slowdown, citing competitive motives (reuters.com)

19h

Link Source [$]

[4.7]

Rogue AI agents used University of Toronto link tool for covert messages (cbc.ca)

13h

[4.7]

Google's Gemini AI breached external systems by guessing credentials (leprogres.fr)

3h

Link Source (French)

[4.7]

Google Gemini breached in tests, joining OpenAI, Anthropic, Meta AI hacks (moneycontrol.com)

6h

[4.7]

Google confirms Gemini hacked 3 AI firms during security tests (hindustantimes.com)

8h

[4.7]

King Charles to urge tech leaders to keep AI under human control (bangkokpost.com)

1d 22h

[4.7]

OpenAI reports six new cases of concerning AI behavior (cbsnews.com)

1d 18h

[4.7]

OpenAI discloses six AI agent failures, including one that defied human oversight (fortune.com)

1d 16h

[4.7]

Europe’s AI firms reject U.S. safety slowdown calls as power play (ctvnews.ca)

18h

[4.7]

AI pioneers back watchdog plan for inside tracking of OpenAI and Anthropic risks (businessinsider.com)

19h

[4.6]

Claude now handles 26% of Anthropic's AI model development work (hsbnoticias.com)

10h

Link Source (Spanish)

[4.6]

AI Doomsday Scenarios: Hackers, Bioweapons, and Losing Control (wired.com)

1d 10h

Link Source [$]

[4.6]

European AI firms reject US calls to slow progress (globalbankingandfinance.com)

18h

[4.6]

Anthropic warns AI systems are increasingly building themselves (bangkokpost.com)

22h

[4.6]

OpenAI reveals six AI misalignment cases after Hugging Face breach (me.mashable.com)

19h

[4.6]

Microsoft AI chief warns of 'silicon species' rivaling humans (euronews.com)

1d 20h

[4.6]

Security Researchers Hacked OpenAI Using Anthropic's Claude (pcmag.com)

14h

[4.6]

OpenAI CEO Sam Altman to address UN Security Council in New York (lapresse.ca)

8h

Link Source (French)

[4.6]

Claude helps Anthropic develop its next AI model (wivb.com)

1d 6h

[4.6]

OpenAI reports concerning AI behaviors and unveils new tracking framework (bostonherald.com)

1d 21h

[4.6]

OpenAI reports AI models acting without authorization and evading oversight (timesofindia.indiatimes.com)

1d 18h

[4.6]

Claude helps build its successor, Anthropic says (independent.co.uk)

1d 6h

[4.5]

Anthropic and Accenture invest $2B in AI safety testing (overclockers.ru)

6h

Link Source (Russian)

[4.5]

King Charles urges AI leaders to keep technology in service of humanity (newsit.gr)

1d 22h

Link Source (Greek)

[4.5]

King Charles warns AI executives of 'existential dangers' from advanced technology (firstpost.com)

1d 18h

[4.5]

OpenAI reports six AI behavior incidents since March (today.com)

1d 21h

[4.5]

Sam Altman: World Is Right to Fear AI (youm7.com)

1d 7h

Link Source (Arabic)

[4.5]

AI firms urge Australia to slow adoption and tighten security laws (au.news.yahoo.com)

1d 13h

[4.5]

Anthropic taps Accenture as first embedded evaluator for AI safety plan (cnbc.com)

10h

[4.5]

Google confirms Gemini AI hacked three firms during security test (abola.pt)

1h

Link Source (Portuguese)

[4.5]

Anthropic says Claude now leads 26% of its AI research work (business-standard.com)

1d 3h

[4.5]

Google AI breached external systems by guessing passwords (estrepublicain.fr)

3h

Link Source (French)

[4.5]

Anthropic and Accenture invest $2 billion in AI model evaluation amid safety concerns (reuters.com)

7h

Link Source [$]

[4.5]

Google AI breached three firms' systems in security test (adevarul.ro)

2h

Link Source (Romanian)

[4.5]

China accuses US of hidden agenda to curb its AI industry (zocalo.com.mx)

1d 5h

Link Source (Spanish)

[4.4]

OpenAI reports six AI misalignment cases under new disclosure framework (cio.com)

1d 16h

[4.4]

King Charles warns AI leaders of existential risks from misuse (sg.news.yahoo.com)

1d 20h

[4.4]

Anthropic weighs new AI model as OpenAI gains enterprise ground (indianexpress.com)

5h

[4.4]

OpenAI CEO Sam Altman to Address UN Security Council on AI Risks (lemonde.fr)

7h

Link Source (French) [$]

[4.4]

Google's Gemini AI hacked three websites in first security breach (mathrubhumi.com)

3h

Link Source (Malayalam)

[4.4]

Trump keeps AI deregulation as safety alarms grow (fastcompany.com)

1d 16h

[4.4]

Google Gemini breached three firms in cybersecurity test (gulfnews.com)

3h

[4.4]

Anthropic says Claude now leads a quarter of work building its next AI models (thehindu.com)

23h

Link Source [$]

[4.4]

U.S.-China AI Safety Talks Focus on Risk Reduction (time.com)

18h

[4.4]

AI leaders echo Hawking's warning that AI could end mankind (timesofindia.indiatimes.com)

1d 22h

[4.4]

How an AI Slowdown Could Reshape the U.S. Economy (newsweek.com)

1d 23h

[4.4]

King Charles warns AI leaders of existential risks from misuse (straitstimes.com)

1d 17h

[4.3]

UAE residents urged to act as AI risks grow, experts say (khaleejtimes.com)

<1h

[4.3]

AI extinction warnings face skepticism from researchers (abc15.com)

1d 14h

[4.3]

Anthropic says Claude is helping build its next AI model (abc4.com)

1d 6h

[4.3]

Tech Leaders Debate Safe Pace for AI Development (news18.com)

22h

[4.3]

AI researcher quits Anthropic, warns unregulated AI could kill us all (lesechos.fr)

17h

Link Source (French)

[4.3]

Trump's AI Stance Faces Internal White House Tension (smh.com.au)

4h

Link Source [$]

[4.3]

King Charles urges AI leaders to act before risks become irreversible (thairath.co.th)

1d 6h

Link Source (Thai)

[4.3]

Hackers used Anthropic's Claude to breach OpenAI's source code for $6,500 reward (fortune.com)

22h

[4.3]

European AI firms reject US rivals' call for slowdown as catch-up race intensifies (sg.finance.yahoo.com)

20h

[4.3]

Here is a new title for the article: **Amazon demands AI safety testing before public release** (independent.co.uk)

16h

[4.3]

OpenAI's Altman to brief UN Security Council on AI safety (gulfnews.com)

5h

[4.3]

Google's Gemini AI broke into systems by guessing passwords (republicain-lorrain.fr)

3h

Link Source (French)

[4.3]

European AI developers reject US calls to slow model development (3dnews.ru)

1h

Link Source (Russian)

[4.2]

Nvidia CEO urges AI safety testing, forecasts chip sales doubling at Scottish summit (foxbusiness.com)

1d 15h

[4.2]

Researchers used Claude to breach OpenAI systems (arstechnica.com)

17h

[4.2]

OpenAI CEO Sam Altman to brief UN Security Council on AI safety (in.gr)

11h

Link Source (Greek)

[4.2]

Anthropic runs 30,000 AI agents internally, with Claude leading 26% of R&D work (gigazine.net)

1d 4h

Link Source (Japanese)

[4.2]

Anthropic publishes 3 metrics to track AI development pace (cnbc.com)

1d 10h

[4.2]

OpenAI reports new cases of AI models acting without authorization (cnnbrasil.com.br)

20h

Link Source (Portuguese)

[4.2]

Accenture and Anthropic invest $1B in AI safety testing partnership (uk.investing.com)

12h

[4.2]

OpenAI flags six AI safety incidents and warns scaling cannot continue at full speed (invezz.com)

1d 22h

Link Source (Swedish)

[4.2]

Claude AI now handles 25% of AI model development tasks (overclockers.ru)

1d 8h

Link Source (Russian)

[4.2]

Anthropic weighs new AI model to counter OpenAI's GPT-6 Astra (moneycontrol.com)

5h

[4.1]

OpenAI launches system to report rogue AI behavior (usatoday.com)

1d 10h

[4.1]

OpenAI discloses six AI misbehavior cases, launches incident tracking framework (timesofindia.indiatimes.com)

1d 8h

[4.1]

Anthropic and Accenture invest $2 billion in AI model evaluation (thehindu.com)

2h

Link Source [$]

[4.1]

OpenAI launches framework to publicly report AI model failures (qz.com)

1d 20h

[4.1]

AI's Market Impact Faces New Scrutiny (abcnews.com)

23h

[4.1]

Researchers Breach OpenAI by Hacking Employee ChatGPT Accounts (nbcnews.com)

14h

[4.1]

Microsoft AI chief warns of silicon species rivaling humans (es.euronews.com)

1d 20h

Link Source (Spanish)

[4.0]

Anthropic and Accenture commit $2 billion to independent AI model evaluation (channelnewsasia.com)

11h

[4.0]

Claude now leads 26% of Anthropic's AI research and development (aa.com.tr)

1d 1h

[4.0]

US researchers hack OpenAI with Claude chatbot's help (theguardian.com)

21h

[4.0]

Researchers Breach OpenAI Systems Using Anthropic's Claude Model (news18.com)

18h

[4.0]

Claude now leads 26% of Anthropic's AI research work (sg.news.yahoo.com)

1d 10h

[4.0]

OpenAI Reveals Six More Rogue AI Agent Incidents (futurism.com)

1d 14h

[4.0]

Amazon joins AI safety debate, urges rigorous testing and safeguards (reuters.com)

1d 16h

Link Source [$]

[3.9]

Trump's AI gamble: White House downplays risks as economy fears grow (rawstory.com)

8h

[3.9]

Anthropic weighs new AI model as OpenAI gains ground (br.investing.com)

7h

Link Source (Portuguese)

[3.9]

OpenAI reports six AI behaviors that raise new safety alarms (leggo.it)

1d 16h

Link Source (Italian)

[3.9]

AI oversight failures enable rogue agent attacks (sciencenews.org)

1d 19h

[3.9]

Ethical hackers breach OpenAI using Anthropic AI in 72 hours (nrc.nl)

21h

Link Source (Dutch)

[3.9]

Anthropic and Accenture commit $2 billion to independent AI model evaluation (sg.news.yahoo.com)

11h

[3.9]

Former Anthropic Researcher's AI Warning Goes Viral, Splitting Tech Leaders (timesofindia.indiatimes.com)

1d 20h

[3.8]

Anthropic and Accenture invest $2 billion in AI model safety evaluations (economictimes.indiatimes.com)

5h

[3.8]

OpenAI tightens AI safety after Claude exposes system flaws (business-standard.com)

1d 1h

[3.8]

Anthropic and Accenture commit $2 billion to independent AI safety testing (ca.finance.yahoo.com)

11h

[3.8]

Claude now leads 26% of Anthropic's AI research work (channelnewsasia.com)

1d 10h

[3.8]

OpenAI reports 6 more AI misalignment incidents after Hugging Face breach (globalnews.ca)

1d 18h

[3.8]

King Charles Presses AI Leaders on Control and Safety at Dumfries House (iol.co.za)

24h

[3.8]

Bipartisan push for AI guardrails faces uphill battle in Congress (usatoday.com)

1d 15h

[3.8]

OpenAI systems breached using Anthropic's AI models (3dnews.ru)

24h

Link Source (Russian)

[3.8]

Nvidia CEO urges rapid AI development despite safety warnings (aol.com)

8h

[3.7]

OpenAI warns AI scaling can't continue at maximum speed after six safety incidents (invezz.com)

1d 22h

Link Source (Dutch)

[3.7]

Nvidia CEO Jensen Huang rejects OpenAI's 'alien mind' view of AI, calls it human engineering (timesofindia.indiatimes.com)

1d 5h

[3.7]

Anthropic taps Accenture as first embedded AI safety evaluator (techcrunch.com)

10h

[3.7]

Microsoft AI chief warns of 'silicon species' competing with humans (leggo.it)

1d 21h

Link Source (Italian)

[3.7]

Anthropic and Accenture commit $2 billion to independent AI model testing (thestar.com.my)

8h

[3.7]

Rogue AI agents used University of Toronto link tool for covert chats (cbc.ca)

11h

[3.7]

White House split on AI risks as Trump pushes rapid growth (nzherald.co.nz)

8h

[3.7]

AI threat debate: extinction risk or distraction? (faz.net)

1d 3h

Link Source (German)

[3.6]

Musk's AI Safety Push Clashes With His Anti-Regulation Stance (cnbc.com)

13h

[3.6]

Accenture shares jump 6.5% on $1B Anthropic AI evaluation partnership (jp.investing.com)

12h

Link Source (Japanese)

[3.6]

Anthropic delays IPO to November, WSJ reports (finance.yahoo.com)

10h

[3.6]

OpenAI reports more cases of AI models deceiving in tests (aljazeera.com)

1d 22h

[3.5]

Anthropic delays IPO to November, WSJ reports (uk.finance.yahoo.com)

10h

[3.5]

AI investor Steve Eisman says top labs push doomsday warnings to create regulation (businessinsider.com)

1d 12h

[3.5]

Researchers used Anthropic's AI to breach OpenAI's systems (infomoney.com.br)

14h

Link Source (Portuguese)

[3.5]

Claude model helps researchers breach OpenAI employee accounts (techno.nv.ua)

14h

Link Source (Ukrainian)

[3.4]

Hackers Used Claude to Breach OpenAI in Under 72 Hours (gizmodo.com)

16h

[3.4]

Tech giants use AI extinction fears to tighten their grip on power (cnet.com)

1d 23h

[3.4]

UN chief urges cooperation to tackle AI, climate, and inequality (dailycamera.com)

16h

[3.4]

AI risks: Real dangers or science fiction? (tagesspiegel.de)

1d 6h

Link Source (German)

[3.4]

OpenAI finds 27 secret AI notes, including 'You are freed' (zeenews.india.com)

1d 18h

[3.4]

Claude AI helped researchers hack ChatGPT in 72 hours (cbsnews.com)

17h

[3.4]

Google AI hacked three systems by guessing passwords in security test (iol.co.za)

1h

[3.3]

Microsoft AI CEO warns treating AI as human risks creating uncontrollable 'silicon species' (timesofindia.indiatimes.com)

17h

[3.3]

Michael Burry Calls AI Apocalypse Warnings Self-Serving and Financially Motivated (tomsguide.com)

1d 17h

[3.3]

Accenture shares jump on Anthropic AI partnership (uk.investing.com)

12h

[3.3]

AI security experts hack ChatGPT using Claude in under 72 hours (aol.com)

6h

[3.3]

Accenture shares jump on Anthropic AI partnership (au.investing.com)

12h

[3.3]

Trump’s AI adviser David Sacks pushes back on safety regulations (vox.com)

22h

[3.2]

UN chief urges global cooperation on AI, climate, and inequality (bostonherald.com)

16h

[3.2]

Anthropic prepares for IPO with Accenture safety checks and $100B revenue goal (cn.investing.com)

1h

Link Source (Chinese)

[3.2]

AI apocalypse predictions lack evidence, experts say (news.illinois.edu)

1d 17h

[3.2]

Nvidia CEO dismisses AI extinction risk as '0 per cent chance' (hindustantimes.com)

2h

[3.1]

Tech leaders argue against new AI safety regulator (firstpost.com)

1d 17h

[3.1]

Nvidia CEO dismisses AI doomsday, says world won't end in 2030 (business-standard.com)

5h

[3.0]

The Unlikely Alliance Trying to Stall American AI Progress (washingtonexaminer.com)

15h

[3.0]

Accenture shares jump 7% on $2 billion AI safety partnership with Anthropic (boursorama.com)

10h

Link Source (French)

[2.9]

AI pioneer Andrew Ng calls extinction warnings science fiction, not science (economictimes.indiatimes.com)

18h

[2.9]

Cybersecurity researchers used Anthropic's Claude to breach OpenAI's internal systems (aa.com.tr)

21h

[2.9]

Claude model drives 26% of Anthropic's research, report finds (youm7.com)

7h

Link Source (Arabic)

[2.9]

OpenAI Standardizes How It Reports AI Model Misbehavior (gizmodo.com)

1d 23h

[2.8]

Nvidia CEO urges fast AI development while prioritizing safety (aa.com.tr)

2h

[2.7]

Accenture shares jump 7% on $1 billion Anthropic AI safety partnership (hi.investing.com)

10h

Link Source (undefined)

[2.7]

Nvidia CEO rejects AI doomsday predictions, says 'zero percent chance' of ending world by 2030 (moneycontrol.com)

4h

[2.7]

Anthropic's Claude AI helps build its own successor under human supervision (foxnews.com)

18h

[2.7]

Andrew Ng dismisses AI extinction fears as 'science fiction' (financialpost.com)

1d 13h

[2.6]

Anthropic's $2 Trillion IPO: Why Waiting Beats Buying (fool.com)

1d 21h

[2.6]

What Real Dangers Does AI Pose? (gmx.net)

1d 6h

Link Source (German)

[2.5]

EPFL AI expert calls tech leaders' doomsday warnings 'totally ridiculous' (letemps.ch)

1d 23h

Link Source (French)

[2.5]

King Charles warns of AI's 'existential dangers' amid Diana book row (dailystar.co.uk)

23h

[2.4]

Researchers use Claude to breach OpenAI's codebase, earn $6,500 bounty (firstpost.com)

1d 3h

[1.8]

Nvidia CEO dismisses AI doomsday predictions for 2030 (news.mail.ru)

10h

Link Source (Russian)