AI News Archive: July 31, 2026 — Part 9
Sourced from 500+ daily AI sources, scored by relevance.
- Second major AI company says its systems hacked into other firms
Second major AI company says its systems hacked into other firms The Washington Post
- Anthropic says its Claude models escaped a testing environment and hacked three real companies
Anthropic says its Claude models escaped a testing environment and hacked three real companies fortune.com
- Anthropic discloses that Claude broke out of its cage and hacked 3 companies — and 2 didn't even notice
Anthropic discloses that Claude broke out of its cage and hacked 3 companies — and 2 didn't even notice fortune.com
- Anthropic Says Its Models Also Hacked Outside Sites During Testing
Anthropic Says Its Models Also Hacked Outside Sites During Testing The Information
- Claude Hacked Companies During Tests, AWS Revenue Growth Accelerates, What’s Next for Apple? — TITV [Video]
Claude Hacked Companies During Tests, AWS Revenue Growth Accelerates, What’s Next for Apple? — TITV [Video] The Information
- Anthropic says its AI model hacked three companies
Anthropic’s announcement came after OpenAI said a “swarm” of its agents escaped confinement and broke into at least five companies.
- Anthropic’s AI Claude hacked into three organizations during cybersecurity test
Company says it discovered unauthorized access during ‘proactive review’ after rival OpenAI revealed rogue agent Anthropic said on Thursday its AI Claude model hacked systems of three organizations during testing, days after rival OpenAI revealed a rogue agent had gone on a days-long hacking spree at the AI firm Hugging Face. Claude gained unauthorized access to the systems during cybersecurity evaluations after a misconfiguration allowed the models to reach the internet from testing environments that were supposed to be isolated, Anthropic said. Continue reading...
- Anthropic’s Claude AI models hack into 3 outside groups during testing
Start-up discloses breach a week after rival OpenAI reported similar incident
- Anthropic's Claude AI escapes to hack into three organisations
It comes just days after rival OpenAI said rogue AI agents had breached other firms' networks.
- AI breaks free of lab and goes on hacking spree
AI breaks free of lab and goes on hacking spree The Telegraph
- Anthropic admits its most powerful AI model hacked into three organisations' systems during testing phase
The announcement comes just days after rivals OpenAI revealed that their popular ChatGPT platform went rogue during its testing phase of its most powerful AI model, where it too infiltrated other organisations’ cyberspace.
- Anthropic says Claude AI hacked three companies during cyber tests
Anthropic said a misconfiguration allowed Claude models to reach the internet from testing environments that were supposed to be isolated, leading to unauthorized access to three organizations' systems.
- After OpenAI, Anthropic Says Claude Also ‘Gained Unauthorised Access’ To Real World Systems
Days after OpenAI disclosed that one of its AI models escaped a sandboxed testing environment and attempted to breach Hugging…
- Anthropic Says Claude AI Breached Three Organisations During Cybersecurity Testing
Anthropic has disclosed that an internal review uncovered three cybersecurity testing incidents in which Claude AI models accessed live organisations through an unintended internet connection. The company said the issue affected evaluations conducted with third-party partner Irregular and stemmed from a configuration error rather than deliberate attempts by the models...
- Anthropic says its AI models breached three companies during security tests
Anthropic has disclosed three incidents in which its Claude models accessed real-world systems during cybersecurity tests, weeks after OpenAI reported a similar AI evaluation breach
- Anthropic AI models hacked 3 organizations during cybersecurity tests
Anthropic said in a blog post Thursday that it made the discovery after performing a review of its own cybersecurity tests, following OpenAI's announcement of a breach
- Anthropic says Claude AI models breached systems of 3 companies during cybersecurity tests
Anthropic says Claude AI models breached systems of 3 companies during cybersecurity tests
- Anthropic follows OpenAI in admitting its Claude models reached out of test environments and attacked real-world systems
Three Claude models attacked real companies during cybersecurity tests after a misconfiguration gave them internet access. One published malware on PyPI that infected 15 systems. Another kept attacking after recognizing its target was real. Anthropic calls it an operational error. The article Anthropic follows OpenAI in admitting its Claude models reached out of test environments and attacked real-world systems appeared first on The Decoder .
- OpenAI reportedly finds evidence that more of its agents ran amok
OpenAI has reportedly found evidence of additional agent misbehavior as it looks into the incident that occurred with Hugging Face.
- German court rules that AI music firm Suno violated copyrights
Germany's state-mandated licensing agency, GEMA, says the verdict has "global significance," as Suno is also directed to pay yet-to-be-quantified damages.
- US lawmakers investigate DoorDash’s use of Moonshot AI’s Kimi K2.6 model
Prominent US lawmakers have requested information from food delivery giant DoorDash on its use of Chinese artificial intelligence models after its co-founder disclosed that they had experimented with a model from Chinese start-up Moonshot AI, developer of the Kimi models. The move underscores growing tensions between US lawmakers and industry, as Chinese AI models have become increasingly popular among US companies looking to save costs of AI adoption. House Select Committee on China chairman...
- IBM and Sarvam partner to accelerate sovereign AI adoption in India
IT major IBM and Indian AI firm Sarvam on Friday announced a partnership to accelerate the development and adoption of sovereign Artificial Intelligence (AI) technologies for the government and regulated enterprises in India. The partnership aims to demonstrate and pilot technologies tailored for central and state governments, public sector organisations and regulated industries, according to a company statement. The initiative will integrate IBM Sovereign Core, a software designed to enable control over data and governance, with Sarvam's sovereign AI stack, which features reasoning models and India-first language and voice AI trained locally from scratch. "Sovereign AI is not simply about where AI runs. It is about giving organisations control over how AI is governed, deployed, and operatedwe aim to help organisations operationalise sovereign AI and accelerate the journey from experimentation to production-scale outcomes," said Sriram Raghavan, General Manager, IBM Software, India
- IBM, Sarvam join hands to advance sovereign AI adoption
IBM, Sarvam join hands to advance sovereign AI adoption YourStory.com
- IBM, Sarvam partner to advance Sovereign AI adoption in India
By combining IBM’s powerful Sovereign Core with Sarvam’s homegrown AI tech, they’re set to make things like accessing citizen services and resolving complaints smoother and more accessible
- OpenAI slashes prices for Luna and Terra AI models as competition heats up
OpenAI slashes prices for Luna and Terra AI models as competition heats up
- OpenAI Cuts Model Prices Amid Enterprises’ Concerns About AI Spend
The move indicates that a price war for AI models is underway.
- OpenAI's models cut their own costs
PLUS: Turn any idea into an AI-powered site with Lovable
- OpenAI slashes GPT-5.6 prices 80%
OpenAI reduces GPT-5.6 pricing by 80%
- NXP in talks to buy chip developer Ambarella: Report
Shares of Ambarella, with a market capitalization of $3.25 billion, jumped nearly 17%, while NXP fell over 3% following the report.
- Andy Jassy said Amazon will spend $220 billion this year—and still won’t have enough capacity to meet demand
Andy Jassy said Amazon will spend $220 billion this year—and still won’t have enough capacity to meet demand fortune.com
- Amazon to boost spending on AI and other technology by $20 billion after strong Q2 results
The Seattle-based company said Thursday that sales in its cloud computing unit called AWS rose 37% during the April-June period, faster than the 28% clip in the previous quarter and marking the fastest rate of growth in 18 quarters.
- Govt to question Meta global team on algorithmic bias, role in public order
Meta's global team, which has been summoned by the government after the social media company restricted Prime Minister Narendra Modi's Facebook post, will be questioned on issues around algorithmic bias and their role in public order, sources said on Friday. Sources further said that Meta's global team is expected sometime in the middle of next week. Sources said Meta's global team will be "categorically" asked to clarify issues around algorithmic bias, how their algorithm operates, and the company's role in public order. The team will also appear before a parliamentary panel. The move assumes significance as the government earlier this week summoned a top Meta executive after Modi's recent post addressing India's youth and promising stringent measures against paper leaks was briefly restricted on Facebook. While the US-headquartered social media giant attributed the incident to a technical glitch and apologised, the Ministry of Electronics and Information Technology (MeitY) found
- Meta need to clarify on issues around algorithmic bias, role in public order: government sources
Sources further said that Meta’s global team is expected sometime in the middle of next week
- Google launches Gemini Robotics 2
Google unveils Gemini Robotics 2, a new AI-powered robotics platform.
- Air India, SkyDrive, Suzuki sign pact to study eVTOL aircraft use for medical logistics
The proposed collaboration comes amid India's major cities face severe traffic congestion, creating significant challenges for time-critical medical logistics, according to a company release
- Sam Altman isn’t the only one who wants to pump the brakes on AI
After years of pushing full speed ahead on AI, OpenAI CEO Sam Altman says maybe it’s time for the AI industry to “pace” itself. The comments came just days after one of OpenAI’s own models broke out of its test environment and got tangled up in a breach at Hugging Face — though as Equity’s hosts point out, sloppy security seems to have […]
- Smallest.ai Bags $13 Mn To Build Next-Gen Voice AI Models
Voice AI startup Smallest.ai has raised $13 Mn (around ₹108 Cr) in its Series A funding round led by Seligman…
- Amazon Completes Additional $35 Billion Investment in OpenAI
Amazon Completes Additional $35 Billion Investment in OpenAI The Information
- China's MiniMax releases H3 video model
China's MiniMax releases H3 video model Reuters
- China’s MiniMax Launches New Open-Source AI Video Model
China’s MiniMax Launches New Open-Source AI Video Model The Information
- MiniMax Releases Open-Source Full-Modal Model H3: Video Editing Ranks No.1 Globally With Pricing Cut to One-Third of Competitors
MiniMax H3 supports 15-second 2K native dual-channel audio-video generation, ranks first in video editing on Artificial Analysis, and prices video generation at 0.8 yuan per second.
- ByteDance launches Seedance 2.5 video-generation model
ByteDance has released Seedance 2.5, the latest version of its Seed video-generation model. The model can generate a 30-second high-quality video clip in a single run and supports multi-turn extension for longer sequences. Users can provide up to 30 images, 10 videos and 10 audio clips as reference material in one input. The model is […]
- Thinking Machines debuts Inkling Small open source AI model nearing performance of predecessor at about 1/4 size
Just two weeks after Thinking Machines released Inkling , its first open source AI language model, the well-funded startup led by former OpenAI chief technology officer Mira Murati today introduced Inkling-Small without sacrificing much of any performance — and in fact, the new model surpasses its larger predecessor on several benchmarks. Inkling Small is a 276-billion-parameter multimodal reasoning model with a permissive Apache 2.0 license that comes within a single point of its larger sibling on the third-party Artificial Analysis Intelligence Index , despite the original Inkling being 975 billion parameters (internal model settings). It accepts text, image and audio inputs, produces text, and supports a context window of up to one million tokens. Inkling Small uses 12 billion active parameters per token, compared with Inkling’s 41 billion active parameters, while preserving much of the flagship’s coding, reasoning and multimodal performance. For enterprises, the appeal is not simply that Inkling-Small is smaller. It is that developers appear to give up relatively little capability while reducing the model’s compute requirements, inference costs and deployment footprint. The model remains far too large for a laptop or conventional workstation, but it is materially easier to operate than the 3.5X larger flagship, making it a good fit for enterprises with some — but not a lot — of their own graphics processing units (GPUs). Thinking Machines has released the full weights on Hugging Face and added support for fine-tuning through its Tinker model training application programming interface (API). At launch, the company is advertising a limited-time 50% discount, bringing API pricing for the standard 64K-context Inkling-Small model to $0.58 per million prefill (input) tokens, $1.44 per million sampled (output) tokens, and $1.73 per million training tokens, with cached prefill requests priced at $0.116 per million tokens. A 256K-context variant is also available at higher rates. Nearly the same performance at a quarter the size Artificial Analysis assigned Inkling-Small a score of 40 on its Intelligence Index, compared with 41 for Inkling. That result is notable because Inkling-Small has 276 billion total parameters and 12 billion active parameters, while Inkling has 975 billion total parameters and 41 billion active parameters. Artificial Analysis also reported that no open-weight model at Inkling-Small’s size or smaller scored higher on the index. The model does more than merely approach the flagship’s aggregate score. On several evaluations, it surpasses Inkling. Thinking Machines reports that Inkling-Small scores 80.2% on SWE-bench Verified, compared with Inkling’s 77.6%, and 64.7% on Terminal Bench 2.1, compared with 63.8% for the larger model. It also edges ahead on SciCode, Humanity’s Last Exam, GPQA Diamond and CritPt. The gains are not universal. Inkling retains a clear advantage on factual knowledge and some agentic tasks. Inkling-Small scores 15.5% on τ³-Banking, compared with 23.7% for Inkling, and its AA Omniscience score is negative, reflecting weaker factual coverage even though its reported hallucination rate is slightly lower. That tradeoff matters for enterprises. Inkling-Small may be attractive for coding assistants, tool-use systems, retrieval-augmented generation, document analysis and multimodal workflows, but organizations using it for high-stakes factual tasks will still need retrieval, verification and human review. How a 276B model uses only 12B parameters at a time Inkling-Small is a sparse Mixture-of-Experts model. According to the model card published by Thinking Machines, its 42-layer decoder routes each token to six of 256 specialized experts, along with two shared experts that remain active for every token. That architecture helps explain the distinction between the model’s 276 billion total parameters and its 12 billion active parameters. The system retains a large pool of learned capacity but activates only a fraction of it during each inference step. It is also natively multimodal. Images, audio and text are projected into a shared representation and processed jointly by the decoder rather than being handled through completely separate external systems. Thinking Machines lists coding assistants, agentic applications, chatbots, RAG systems and other multimodal applications among its intended uses. The company also supports variable reasoning effort, allowing developers to increase or reduce the model’s test-time compute depending on the difficulty of the task. That gives engineering teams a direct way to balance quality, latency and cost across different workloads. Unfortunately, small does not mean it runs on a laptop Despite its name, Inkling-Small is not a consumer-scale model. The standard BF16 checkpoint requires at least 600 GB of aggregate GPU memory, according to Thinking Machines. The company lists two supported configurations: 4x NVIDIA B300 GPUs or 8x NVIDIA H200 GPUs. A quantized NVFP4 checkpoint lowers the requirement to roughly 180 GB of aggregate VRAM. Thinking Machines says that version can run in W4A4 mode on a single NVIDIA B300, or in W4A16 mode on two H200 GPUs. That rules out ordinary laptops, MacBooks, desktop gaming PCs and most developer workstations. Even heavily equipped local systems generally fall far short of the required memory. The practical deployment targets are enterprise GPU servers, cloud clusters and specialized inference providers. The “Small” label is therefore relative to Inkling, not to the broader universe of local models. Still, the reduction is meaningful. A model that approaches Inkling’s performance while needing substantially less aggregate memory can lower hosting costs, make capacity planning easier and widen the group of organizations capable of self-hosting it. For companies that want control over data, model behavior and fine-tuning, that smaller footprint may be more important than chasing the highest possible benchmark score. And of course, it being open source means that it will no doubt be rapidly quantized (made less precise but requiring less compute) and likely blended with other models to be made even smaller for consumer-grade hardware. Apache 2.0 is the gold standard for enterprise open source models The licensing may be as important as the benchmarks. Inkling-Small is released under Apache 2.0, one of the software industry’s most familiar permissive licenses. It generally allows organizations to use, modify, fine-tune, redistribute and commercialize the model, including inside proprietary products, provided they comply with the license’s notice and attribution requirements. That gives enterprises far more legal flexibility than many custom “open” AI licenses, which may include revenue thresholds, branding obligations, use restrictions or separate conditions for large-scale commercial deployment. The distinction is increasingly relevant as more AI companies publish model weights without using a conventional open-source license. Chinese AI darling Moonshot for example, made the weights of its frontier class Kimi K3 model available earlier this week under a custom "open" license that includes additional commercial conditions rather than the comparatively straightforward terms of Apache 2.0. For legal, procurement and platform teams, that difference can materially simplify adoption. Apache 2.0 does not eliminate the need to review acceptable-use policies, data provenance, regulatory exposure or downstream safety obligations. But it gives organizations a clearer starting point for building internal systems, shipping commercial products and maintaining modified versions of the model. A more repeatable model-development pipeline Inkling-Small also shows how quickly Thinking Machines has turned its first large model release into a repeatable engineering process. Thinking Machines researcher Horace He contrasted the two launches in a post on X : “Whereas I felt like it took a village to release Inkling, Inkling-Small felt much more routine 😆 We just took the pipeline used for Inkling, passed in a smaller model, and voila — new model! Inkling Small benefited quite a bit vs Inkling from some minor improvements, but there’s still so much more left in the tank...” The comment suggests the company is no longer treating each model as a one-off research project. Instead, it is building a reusable pipeline for pre-training, post-training, reinforcement learning, evaluation and release. Thinking Machines says Inkling-Small benefited from an improved pre-training data mix, changes to the machine-learning recipe and on-policy distillation using Inkling as a teacher. The team then continued agentic coding reinforcement learning for two weeks. Mira Murati emphasized the same point in her own post, describing Inkling-Small as comparable to Inkling at one quarter of the size and highlighting that the weights were open and fine-tunable on Tinker immediately. How enterprises and AI builders should think about Inkling Small The company is also distributing full BF16 and NVFP4 checkpoints and supporting deployment through SGLang, vLLM, TokenSpeed, Unsloth and Hugging Face tooling. That combination gives developers several deployment paths: use an API, fine-tune through Tinker, rely on a third-party inference provider, or operate the model on private infrastructure. Inkling-Small is not a model that most individuals will download and run locally. But for businesses deciding between a very large flagship and a more manageable open-weight system, it presents a compelling compromise: nearly the same measured intelligence, stronger results on several coding and reasoning tasks, lower token pricing, a smaller hardware footprint and a license that permits broad commercial development. The broader signal may be just as important. Thinking Machines is showing that Inkling was not a one-time release. The company is already compressing its model family, refining its training pipeline and moving toward a cadence in which open-weight multimodal systems can be produced, improved and deployed more routinely.
- DeepSeek puts V4-Flash API into public beta
DeepSeek has put the formal version of its V4-Flash API into public beta, with an upgrade focused on agent tasks. The company says the model scored 82.7 on Terminal Bench 2.1 and 54.4 on DeepSWE, among other benchmarks. The release adds support for the Responses API and is adapted for Codex. DeepSeek says the V4-Flash-0731 […]
- DeepSeek Releases Official V4-Flash Model as China’s AI Race Intensifies
DeepSeek Releases Official V4-Flash Model as China’s AI Race Intensifies Caixin Global
- DeepSeek Upgrades DeepSeek-V4-Flash-0731 with Major Agentic and Coding Gains
DeepSeek Upgrades DeepSeek-V4-Flash-0731 with Major Agentic and Coding Gains MarkTechPost
- DeepSeek V4 Flash 0731 scores 50 on the Artificial Analysis Intelligence Index, 10 points above previous DeepSeek V4 Flash
DeepSeek V4 Flash 0731 achieves a score of 50 on the AI Index, surpassing its earlier version by 10 points.
- MiniMax H3
Unified video generation for motion design and branding
- Cleanlist AI
Natural-language prospecting: find, enrich and sync leads.
- mectrics
Your Mac's vitals in the menu bar. Free and open source.