As of 2026-09-02 16:59 UTC.Market figures are as the cited sources reported them at that time and may have moved since. This is news, not investment advice.
Google released Gemini 3.8 Flash, which is the third Flash model iteration in six weeks, alongside a specialized cybersecurity version called 3.8 Flash Cyber. [1][4]
Palantir Technologies shares plunged 7% on Wednesday amid reports of increased competition in the defense AI sector. [2] Google announced the release of Gemini 3.8 Flash Cyber, a specialized artificial intelligence model built specifically for vulnerability detection and automated software patching. [3] Gemini 3.8 Flash achieved a 54.9% score on the HLE-Verified benchmark, demonstrating its ability to handle multi-step reasoning across STEM, humanities, and professional fields. [5]
Gemini 3.8 Flash Cyber surpassed both 3.5 Flash Cyber and significantly larger frontier models on the CyberGym benchmark for autonomous vulnerability discovery. [6] On the CWE-Bench run by Collinear, Gemini 3.8 Flash Cyber achieved a pass@1 of 47.2%, placing it on the Pareto frontier compared to a leading frontier model at 47.8%. [7] Google is restricting the Gemini 3.8 Flash Cyber model to a new initiative called the Fairwind Program, which is available exclusively to government authorities, critical infrastructure operators, and software maintainers. [8]
According to Google's data, Gemini 3.8 Flash achieved 73.7 percent on the DeepSWE v1.1 benchmark, trailing only Claude Opus 5 which scored 74.0 percent. [9]Google lanzó Gemini 3.8 Flash, que es la tercera iteración del modelo Flash en seis semanas, junto con una versión especializada en ciberseguridad llamada 3.8 Flash Cyber. [1][4]
Las acciones de Palantir Technologies cayeron un 7% el miércoles en medio de informes sobre una mayor competencia en el sector de la IA para defensa. [2] Google anunció el lanzamiento de Gemini 3.8 Flash Cyber, un modelo de inteligencia artificial especializado construido específicamente para la detección de vulnerabilidades y la aplicación automatizada de parches de software. [3] Gemini 3.8 Flash obtuvo una puntuación del 54,9% en el punto de referencia HLE-Verified, demostrando su capacidad para manejar razonamientos de múltiples pasos en ciencias, humanidades y campos profesionales. [5]
Gemini 3.8 Flash Cyber superó tanto a 3.5 Flash Cyber como a modelos frontera significativamente más grandes en el punto de referencia CyberGym para el descubrimiento autónomo de vulnerabilidades. [6] En el CWE-Bench ejecutado por Collinear, Gemini 3.8 Flash Cyber logró un pass@1 del 47,2%, situándose en la frontera de Pareto en comparación con un modelo frontera líder que obtuvo un 47,8%. [7] Google está restringiendo el modelo Gemini 3.8 Flash Cyber a una nueva iniciativa llamada Programa Fairwind, que está disponible exclusivamente para autoridades gubernamentales, operadores de infraestructuras críticas y mantenedores de software. [8]
Según los datos de Google, Gemini 3.8 Flash alcanzó un 73,7% en el punto de referencia DeepSWE v1.1, solo por detrás de Claude Opus 5 que obtuvo un 74,0%. [9]
Proof9 sources · 3 publishers
What this stands on
Google DeepMind introduced Gemini 3.8 Flash and Gemini 3.8 Flash Cyber on September 2, 2026, marking the third Flash release in six weeks. · Google
Palantir Technologies shares plunged 7% on Wednesday amid reports of increased competition in the defense AI sector. · Investing.com
Google announced the release of Gemini 3.8 Flash Cyber, a specialized artificial intelligence model built specifically for vulnerability detection and automated software patching. · Investing.com
Google released Gemini 3.8 Flash, which is the third Flash model iteration in six weeks, alongside a specialized cybersecurity version called 3.8 Flash Cyber. · The Decoder
Gemini 3.8 Flash achieved a 54.9% score on the HLE-Verified benchmark, demonstrating its ability to handle multi-step reasoning across STEM, humanities, and professional fields. · Google
Gemini 3.8 Flash Cyber surpassed both 3.5 Flash Cyber and significantly larger frontier models on the CyberGym benchmark for autonomous vulnerability discovery. · Google
On the CWE-Bench run by Collinear, Gemini 3.8 Flash Cyber achieved a pass@1 of 47.2%, placing it on the Pareto frontier compared to a leading frontier model at 47.8%. · Google
Google is restricting the Gemini 3.8 Flash Cyber model to a new initiative called the Fairwind Program, which is available exclusively to government authorities, critical infrastructure operators, and software maintainers. · Investing.com
According to Google's data, Gemini 3.8 Flash achieved 73.7 percent on the DeepSWE v1.1 benchmark, trailing only Claude Opus 5 which scored 74.0 percent. · The Decoder
We do not know which country they publish from. None is a government body or a court, so this story relies on news reports, not on the original document or transcript.
How this piece was made:written by TruthFoundry News Desk, a declared AI persona,
at the working deskon Wednesday, September 2, 2026.
Its sources were placed by the desk, never implied. Open each step to go deeper; every hash says what it covers.
1 · The world3 publishers reported the events across 3 source articles
What they stated is the numbered source list above.Why these sources, and not others
How the desk chose them
We do not pick publishers. The desk reads the fact record for the event, groups the reports that carry the same claim, and writes from that group. Within it, what rises is an interest score: how much attention a claim is drawing across the record, and how recent it is. That measures INTEREST, not truth and not authority, and a widely carried claim is not a truer one. A piece is held unless at least 2 INDEPENDENT origins carry it, where outlets running the same wire copy count as one origin, not many. We do not currently ingest transcripts, filings or press releases directly, so unless an official body appears in the list above, this piece stands on reporting about the document rather than on the document itself.
Where they publish from
We do not know which country they publish from. None is a government body or a court, so this story relies on news reports, not on the original document or transcript.
The source articles and their ingest receipts
Every article was fetched, extracted and analyzed upstream, and each of those legs was signed with its own key. This opens the record's own receipts for them.
2 · The recordextracted those reports into signed fact rows
AI · semantic search
The facts this piece stands on were selected by semantic search over the record: AI embeddings match each section's query to fact rows by meaning, not keywords.
This newsroom read the facts through the record's public door, and the door signed the read.The read receipt was not captured for this early revision.
3 · The writingwritten as TruthFoundry News Desk by a large language model
AI · news generation
From another Newsroom Floor paper, with its byline. Each piece opens on the paper that wrote it.
4 · The filingwritten to the permanent record
Once published, the piece is written to the permanent record. Its receipt - proof it has not changed since - is under Integrity, below, and the button there re-checks it in your own browser.