Artificial Intelligence · Cybersecurity
SentinelOne expands Wayfinder with OpenAI models
SentinelOne said it had expanded Wayfinder Frontier AI Services with OpenAI Daybreak models, plus code-risk and compromise-assessment functions, starting with GPT-5.6-Cyber and rolling the additions out in private preview.
- On 3 September 2026 in Mountain View, SentinelOne said it had expanded Wayfinder Frontier AI Services.
- The company said the service would use OpenAI Daybreak models through the Daybreak Defense Network, beginning with GPT-5.6-Cyber.
- SentinelOne said its SentinelLABS team had found GPT-5.6-Cyber best in its testing for reverse engineering and malware analysis, including work on a sample called fast16.
- The update added AI-powered code risk analysis and AI-enabled compromise assessment, with broader availability to follow after private preview.
SentinelOne announced an expansion of Wayfinder Frontier AI Services on 3 September 2026. It said the package would add code-risk analysis, compromise assessment and malware analysis powered by OpenAI Daybreak models.
The company said the first model in the set would be GPT-5.6-Cyber. It said the models were being used through OpenAI’s Daybreak Defense Network.
SentinelOne said its AI security and research teams had tested public and private models for cyber tasks. In recent benchmarks, its SentinelLABS unit said GPT-5.6-Cyber performed best in reverse engineering and malware analysis, including analysis of a sample it called fast16.
The company said the new functions included scans for flaws, code implants, exposed secrets and supply-chain risk. It also said compromise assessment would use AI-assisted triage to review telemetry for risky tools and other indicators.
SentinelOne said the new capabilities were rolling out in private preview, with wider availability planned later. Steve Stone, the company’s chief customer officer, said the work was intended to help customers prioritise what an attacker could reach and how to respond.
Named in this story
People
- Steve Stone
- SentinelOne’s chief customer officer who commented on the announcement
Companies
- SentinelOne
- announced the expanded Wayfinder Frontier AI Services offering
- OpenAI
- supplied the Daybreak models used in the offering
Organisations
- Daybreak Defense Network
- the network through which the OpenAI models were provided
- SentinelLABS
- SentinelOne’s research arm that ran the benchmarks
Products and systems
- Wayfinder Frontier AI Services
- the service line that was expanded
- GPT-5.6-Cyber
- the OpenAI model SentinelOne said it would use and benchmarked
- fast16
- the malware sample cited in SentinelLABS testing
Places
- Mountain View, Calif.
- the place from which SentinelOne announced the update
How the source tells it
The piece read as a promotional company announcement, with repeated performance claims and threat-imminence language used to frame the product update.
- vendor boosterism The release repeatedly promoted SentinelOne’s own leadership, expert teams and customer value as settled fact.
- hype It leaned on frontier-model language and benchmark claims to present the update as highly advanced.
- urgency It framed exposures and attacker activity as matters that had to be identified and closed quickly.