595 summarised stories about OpenAI, each linking back to the original source. Browse all topics →
OpenAI 1 year ago 35
OpenAI released a Realtime API that enables developers to build applications with fast speech-to-speech capabilities. The API supports bidirectional audio streaming with response times measured in hundreds of milliseconds. This allows developers to create interactive voice applications with lower latency than previous API options.
OpenAI 1 year ago 27 ● 4 sources
OpenAI terminated accounts linked to SweetSpecter, a suspected China-based cyber group, which had been using the platform's AI to research security vulnerabilities, generate malicious code, and craft spear-phishing messages. The group's activities were detected and documented by OpenAI's threat intelligence team working with external partners. Removing these accounts reduces the attacker's access to AI-assisted cyber operations, though similar groups may simply migrate to other AI platforms.
OpenAI 1 year ago 41 ● 4 sources
OpenAI shut down accounts linked to CyberAv3ngers, an Iran-connected group that was using AI to research industrial control systems and default credentials. The activity was detected and the accounts were terminated before any confirmed attacks occurred. The action reflects how AI platforms are now monitoring for state-sponsored cyber reconnaissance using their tools.
OpenAI 1 year ago 23
OpenAI suspended an account that used AI to generate a fake ChatGPT error message falsely claiming to detect Russian troll activity. The hoax message mimicked ChatGPT's interface to spread disinformation. The incident highlights how AI tools can be misused to create convincing false error messages and spread unverified claims.
OpenAI 1 year ago 6 ● 2 sources
OpenAI removed accounts engaged in a coordinated inauthentic behavior campaign called "A2Z" that used AI to generate multilingual content promoting political narratives about elections, Ukraine, and political issues across social media platforms. The operation targeted multiple languages and coordinated across at least several platforms to spread influence content at scale. The removal disrupts this particular network's ability to use OpenAI's tools for coordinated inauthentic amplification of political messaging.
OpenAI 1 year ago 40
OpenAI terminated accounts that an Israel-based startup used to generate spam conversations and distribute gambling site links on X using OpenAI's models. The network involved accounts accessing OpenAI's API to create automated posts promoting betting platforms at scale. The action prevents further abuse of OpenAI's infrastructure for coordinated gambling promotion across social media.
OpenAI 1 year ago 13 ● 3 sources
OpenAI removed accounts originating from Rwanda that were using AI tools to generate partisan political comments before the country's elections. The removals targeted coordinated inauthentic behavior using OpenAI's models to create election-related content at scale. The action prevents AI-generated disinformation campaigns from influencing voters during a sensitive political period.
OpenAI 1 year ago 9 ● 3 sources
OpenAI suspended accounts that were using its AI tools to generate abusive reports and complaints targeting Vietnamese public figures and online platforms. The company did not disclose the number of accounts removed or the scale of the campaign. This action reflects OpenAI's enforcement of policies against using AI systems for harassment and coordinated abuse.
OpenAI 1 year ago 26 ● 3 sources
OpenAI banned accounts that were using AI to generate comments criticizing a Russian anti-corruption foundation and related figures. The accounts were caught using the platform's models to create inauthentic critical content at scale. This represents enforcement of OpenAI's policy against coordinated inauthentic behavior and misuse of AI for disinformation campaigns.
The daily briefing
TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.
TLDRocket
8 August 2026
Mistral AI's release of Shieldstral 1.0 3B marks a practical inflection point in content moderation: safety classifiers are shedding their fixed rulebooks. The 3B model takes policies as plain-language prompts at inference time—meaning a single deployment can enforce different standards for different customers without retraining—while matching the performance of a 20B baseline on both text and multimodal content. It runs on a single 16GB GPU and ships under Apache 2.0, making it available for local deployment across vLLM and llama.cpp. For teams managing multiple content policies or operating across jurisdictions, this matters because moderation no longer requires separate model forks or constant retraining cycles.
Read the full briefing →