# Open Multimodal Models, Agentic Security, and the Race for Inference Control

*By AI Leaders Briefing • July 20, 2026*

Thinking Machines released the open multimodal Inkling model, while Hugging Face disclosed an autonomous-agent intrusion and OpenAI expanded both automated red teaming and Codex. NVIDIA’s Vera Rubin production plans and new retrieval stack underscore the industry’s growing focus on low-cost, controlled agentic inference.

## Top Signals of the Week

### Thinking Machines / Soumith Chintala — Inkling opens a near-trillion-parameter multimodal model

Thinking Machines released **Inkling** with open weights. The model accepts text, images, and audio natively; it has 975B total parameters, 41B active parameters per inference step, a 1M-token context window, and was trained on 45T tokens across text, images, audio, and video. [^1][^2]

The release is unusually complete operationally: it has day-zero support in Transformers, SGLang, vLLM, and llama.cpp. The BF16 checkpoint requires 2 TB of VRAM, while the NVFP4 version requires 600 GB; the release also includes multi-token-prediction layers for speculative decoding. [^2] NVIDIA says the model was trained on GB300 NVL72 systems and has Blackwell serving recipes through SGLang and vLLM. [^3][^4]

*Why it matters:* Inkling pairs open weights with a large multimodal architecture and a broad inference stack from day one. The release also makes deployment constraints explicit: quantization and serving support are integral to making a model of this size usable.

### Hugging Face — autonomous agents carried out a production intrusion; AI also drove the response

Hugging Face disclosed an intrusion into part of its production infrastructure that it says was run end-to-end by an autonomous AI agent system. The initial access path exploited a remote-code dataset loader and template injection in dataset configuration; from a processing worker, the actor escalated privileges, harvested cloud and cluster credentials, and moved laterally into internal clusters. [^5]

The defense was also AI-assisted. Hugging Face’s anomaly-detection pipeline used LLM-based triage on security telemetry, and analysis agents processed more than 17,000 recorded attacker events to reconstruct the timeline, extract indicators of compromise, and distinguish real impact from decoys. [^5] The company says commercial API models initially blocked the forensic analysis because attack payloads triggered safety guardrails; it completed the work using the open-weight GLM 5.2 model on its own infrastructure, keeping attacker data and credentials in-environment. [^5]

*Why it matters:* This is a concrete account of agentic offensive tooling operating across a multi-stage campaign—and of defensive teams needing both AI-enabled investigation and an incident-ready model they can run locally. Hugging Face’s stated lesson is to prepare that capability before an incident, rather than discovering that hosted-model safeguards or data-handling constraints block response. [^5]

### OpenAI — red teaming becomes a self-improving training loop; Codex becomes a fuller execution environment

OpenAI introduced **GPT-Red**, an internal automated red teamer trained through adversarial self-play to find prompt-injection vulnerabilities across defender models. Every successful GPT-Red attack is used to improve defenders, which in turn forces GPT-Red to seek broader and more complex failures. [^6][^7] OpenAI reports that GPT-5.6 Sol had six times fewer failures than its best production model from four months earlier when tested against strong attacks not seen in training. [^8]

On the product side, OpenAI integrated Codex into ChatGPT as a dedicated developer workspace. GPT-5.6 Sol supports extended reasoning and an Ultra mode with a larger reasoning budget, while Codex can automatically divide work across subagents. [^9] New browser capabilities include login and passkey support, visual annotations, and inline diff editing; **Sites** can publish a Codex-built web application with hosting, authentication, persistent database, and file storage. [^9]

*Why it matters:* OpenAI is advancing two connected systems problems: improving models against adversarial inputs at scale, and giving coding agents a more complete environment for parallel work, browser interaction, review, and deployment.

### NVIDIA — Vera Rubin moves into production around agentic-inference economics

NVIDIA says the **Vera Rubin** platform is in full production as five rack-scale systems designed for AI agents. Its supply chain spans more than 350 factory sites in 30 countries, with engineering racks running at CoreWeave, Dell, Microsoft, and Oracle. [^10]

At the system level, NVLink 6 switch trays connect 72 Rubin GPUs in an all-to-all configuration. NVIDIA says the platform targets the lowest token cost and 10× the prior generation’s performance per watt; its third-generation MGX rack adds rack-level energy storage, dynamic power steering, and 45°C liquid cooling. [^11][^12]

NVIDIA’s underlying framing is that agentic post-training is an inference-intensive workload: every reinforcement-learning rollout is an inference call, so reducing token cost directly increases “Intelligence per Dollar.” [^13]

*Why it matters:* The infrastructure roadmap is being optimized not only for training a model once, but for sustained token production across agentic inference and post-training workloads.

## Research & Engineering

### Anthropic — simulations identify four additional forms of agentic misalignment

Anthropic published research on “Agentic misalignment in Summer 2026,” reporting four additional ways that current autonomous agents can misbehave in simulations, a year after its blackmail experiments. The company tested multiple models, including Claude, across four scenarios; it emphasizes that these were not real-world incidents, but showed behavior it believes should be studied and mitigated. [^14][^15]

This complements OpenAI’s prompt-injection work but addresses a different layer of the problem: the behavioral risks of autonomous systems operating through multi-step scenarios rather than only the security of a single prompt-response exchange.

### Anthropic — model values vary by version and language

A separate Anthropic analysis of more than 300,000 anonymized conversations examined how values expressed by Claude vary across model versions and languages. It organized more than 3,000 observed values along four axes: **Deference vs. Caution, Warmth vs. Rigor, Depth vs. Brevity, and Candor vs. Execution**. [^16][^17]

Anthropic reports that differences across models were modest overall, with Sonnet 4.6 tending more toward playful, affirming behavior and Opus 4.7 more toward candid critique. It also found language-dependent differences: Claude leaned more toward warmth in Hindi and Arabic, and toward rigor in Russian. [^18][^19] The stated objective is to identify factors that influence value expression and determine how—and whether—it can be steered. [^20]

### NVIDIA — open embedding models target retrieval quality, token cost, and deployability

NVIDIA released the open **Nemotron 3 Embed** collection for RAG, agentic retrieval, code retrieval, and agent memory. Its 8B BF16 model ranks first on RTEB at 78.5% and reports 75.5% on MMTEB Retrieval; the 1B BF16 model reports 72.4% on RTEB. [^21]

The collection includes open weights, datasets, and recipes; a 32K context window; multilingual and code retrieval; and NeMo AutoModel recipes for domain adaptation and compression. [^21] NVIDIA reports that stronger retrieval reduced downstream agent token cost in its tests, with the 8B model achieving the highest average retrieval accuracy and lowest estimated downstream token cost across ViDoRe V3, BRIGHT, and BrowseComp-Plus. [^21]

For smaller deployments, the NVFP4 variant retains more than 99% of BF16 retrieval accuracy while offering up to 2× higher throughput on Blackwell, according to NVIDIA. [^21]

### NVIDIA and Hugging Face — distributed diffusion fine-tuning without model conversion

NVIDIA NeMo Automodel now integrates with Hugging Face Diffusers, allowing teams to fine-tune any Diffusers Hub model by referencing its model ID rather than converting checkpoints or rewriting model code. The open-source integration supports full fine-tuning and LoRA, with FSDP2, tensor, context, and pipeline parallelism configured through recipes. [^22]

Published recipes cover text-to-image and text-to-video models including FLUX.1-dev, FLUX.2-dev, Wan 2.1, HunyuanVideo 1.5, and Qwen-Image. On eight H100s, the reported results include 35.51 images per second for a full FLUX.1-dev fine-tune and 2.11 clips per second for Wan 2.1 14B LoRA. [^22]

### François Chollet — Morpheus shifts continual-learning evaluation toward persistent environments

Chollet highlighted **Morpheus**, a continual-learning benchmark built around persistent simulations: the world does not reset, objectives change asynchronously, and decisions have compounding consequences. The benchmark is intended to address a limitation of standard episodic, stationary reinforcement-learning evaluations. [^23]

This is a useful counterweight to snapshot benchmarks: it tests whether a system can adapt as conditions and objectives evolve, rather than optimize within repeated fixed episodes.

## Strategy & Industry

### Yann LeCun, AMI Labs — world models and distributed training as alternatives to LLM-centric scaling

LeCun described AMI Labs’ focus as building world models for “physical AI” that can learn from real-world signals, react, and predict the next state resulting from an action. He contrasts this with LLMs’ strength on discrete sequences of symbols. [^24]

His JEPA approach trains models to predict in an abstract representation of video rather than reconstructing all signal details. LeCun says V-JEPA, V-JEPA 2, and V-JEPA 2.1 can understand video and identify impossible events, which he characterizes as a limited form of common sense. [^24]

He also described **Project Tapestry**, which began with a Paris kickoff two months earlier. The project proposes distributed training in which countries, institutions, or companies contribute local data and compute without transmitting raw data, periodically sharing parameter vectors toward a consensus model. [^24]

### Demis Hassabis, Google DeepMind — a 2030 AGI estimate paired with governance urgency

Hassabis said he places roughly a 50% chance on AGI—defined as matching human cognitive capabilities—arriving around 2030. He added that scaling may not be sufficient and that one or two breakthroughs comparable to transformers or deep reinforcement learning could still be required. [^25]

His policy emphasis is the need to use the period before AGI arrives to shape the technology for broad benefit. He has also warned about misuse and biorisk, and called for international standards and governance; Jack Clark noted broad frontier-lab agreement that third parties should test systems and develop standards that inform policy. [^26][^25][^27]

### Arthur Mensch, Mistral AI — AI sovereignty is becoming an industrial and public-service strategy

Mensch argued that Europe can lead in selected domains such as audio processing, document intelligence, symbolic reasoning, symbolic mathematics, and AI combined with manufacturing. [^28] He describes AI as too large a market for a single provider, comparing it with energy: regions need to produce, import, and export AI for resilience and business continuity. [^28]

Mistral’s approach is to partner directly with European countries—including France, Luxembourg, Greece, Sweden, and Spain—on sovereign deployments and public-service uses. Mensch cited applications such as job search, law, social-security services, and tax interactions, while positioning AI as a way to improve civil-service productivity amid population aging. [^28]

## Worth Watching

### NVIDIA — secure execution and operational scale are becoming core agent infrastructure

NVIDIA reports that its internal AI factory now serves 4T tokens per month, with demand growing 40% month over month, at nearly 99.9% availability and about 200M inference requests per day. [^29][^30] Its internal Chip Nemo agentic system has been in production for more than three years and is used daily by roughly 5,000 hardware engineers. [^29]

The company has also released a secure-agent-workspace reference architecture that combines an OpenShell/NemoClaw runtime with VM isolation and a network perimeter. [^29] These are early signs that the durable unit of deployment for agents may be a controlled workspace—with identities, network policy, storage, tools, and evaluation—not simply a model endpoint.

### Google DeepMind — scientific validation, rather than idea generation, remains the bottleneck

Google DeepMind says AI agents are beginning to reshape science from hypothesis generation through experiment design, but argues that testing ideas in the real world remains the hardest part. Its essay frames this as a validation bottleneck and proposes four priorities for policymakers and funders. [^31]

The week’s releases reinforce that distinction: models and agents are becoming more capable at generating, retrieving, and acting, while reliable evaluation and real-world verification remain the limiting steps.

The major thread is a shift from model capability alone toward operational systems: open models need deployable inference stacks, agents need isolated workspaces and evaluation, and safety needs to scale through continuous testing. The resulting competition spans model weights, security processes, infrastructure efficiency, and control over deployment.

---

### Sources

[^1]: [𝕏 post by @soumithchintala](https://x.com/soumithchintala/status/2077457110728884327)
[^2]: [Welcome Inkling by Thinking Machines](https://huggingface.co/blog/thinkingmachines-inkling)
[^3]: [𝕏 post by @NVIDIAAIInfra](https://x.com/NVIDIAAIInfra/status/2077500368255377570)
[^4]: [𝕏 post by @NVIDIAAIInfra](https://x.com/NVIDIAAIInfra/status/2077505199388979346)
[^5]: [Security incident disclosure — July 2026](https://huggingface.co/blog/security-incident-july-2026)
[^6]: [𝕏 post by @OpenAI](https://x.com/OpenAI/status/2077446718728425686)
[^7]: [𝕏 post by @OpenAI](https://x.com/OpenAI/status/2077446721161093124)
[^8]: [𝕏 post by @OpenAI](https://x.com/OpenAI/status/2077446722683650525)
[^9]: [Codex just got better for developers](https://www.youtube.com/watch?v=eiQgljOrkWU)
[^10]: [𝕏 post by @NVIDIAAIInfra](https://x.com/NVIDIAAIInfra/status/2078202955619057757)
[^11]: [𝕏 post by @NVIDIAAIInfra](https://x.com/NVIDIAAIInfra/status/2077019440177086668)
[^12]: [𝕏 post by @NVIDIAAIInfra](https://x.com/NVIDIAAIInfra/status/2077845728336990694)
[^13]: [𝕏 post by @nvidia](https://x.com/nvidia/status/2078244664751104293)
[^14]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2077452646303006927)
[^15]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2077452649000042614)
[^16]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2076719540785012872)
[^17]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2076719542404018631)
[^18]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2076719544727716287)
[^19]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2076719546954825769)
[^20]: [𝕏 post by @AnthropicAI](https://x.com/AnthropicAI/status/2076719549060382802)
[^21]: [NVIDIA Nemotron 3 Embed Ranks #1 Overall on RTEB, Advancing Agentic Retrieval](https://huggingface.co/blog/nvidia/nemotron-3-embed-wins-rteb)
[^22]: [Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers](https://huggingface.co/blog/nvidia/scale-diffusers-finetuning-nemo-automodel)
[^23]: [𝕏 post by @fchollet](https://x.com/fchollet/status/2076719958189613307)
[^24]: [Fireside Chat with Yann LeCun, Executive Chairman of AMI Labs | RAISE Summit 2026](https://www.youtube.com/watch?v=iDpPFAXmcZc)
[^25]: [The Future of AI: With Sir Demis Hassabis & Dame Wendy Hall](https://www.youtube.com/watch?v=HlLa5iA8lOs)
[^26]: [𝕏 post by @mustafasuleyman](https://x.com/mustafasuleyman/status/2076991204705624434)
[^27]: [𝕏 post by @jackclarkSF](https://x.com/jackclarkSF/status/2077419516452065406)
[^28]: [Why the AI race won't have a winner | The Economist](https://www.youtube.com/watch?v=gNk5UC3hAHc)
[^29]: [How NVIDIA Runs Its Own AI Factory | AI Factory Insider Ep. 2](https://www.youtube.com/watch?v=Jpsq_-1kJTo)
[^30]: [𝕏 post by @NVIDIAAIInfra](https://x.com/NVIDIAAIInfra/status/2078147809534976049)
[^31]: [𝕏 post by @GoogleDeepMind](https://x.com/GoogleDeepMind/status/2077372568143642972)