Improved controls for the new era of cyberattacks
Blog post from Baseten
The rapid improvement of AI models and agent frameworks is enabling faster, increasingly autonomous cyberattacks, according to Baseten, with risks spanning software supply chains, data exfiltration, and access to model weights and computing resources. The company argues that faster software releases and expanding attack surfaces make older security practices inadequate, and describes using layered infrastructure controls, AI-assisted monitoring and response, and agentic red teaming to identify and address threats. For customers, Baseten highlights identity-provider integration and fine-grained roles, short-lived workload tokens, scoped and revocable API keys, customer-managed encryption keys, configurable network boundaries, approved inference regions, zero data retention, and exportable audit logs. It presents security as an ongoing effort and says it plans to share more about its practices for cluster security, credentials, containers, identity, and incident detection and response.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| AI Guardrails | 1 | 253 | 105 | 43 | -54% |
| LLM | 1 | 3,674 | 823 | 199 | -44% |
| Multi-agent systems | 1 | 173 | 95 | 52 | -47% |
| Observability | 1 | 2,390 | 523 | 156 | -23% |
| Platform Engineering | 1 | 1,295 | 290 | 84 | -36% |
| Secrets Management | 1 | 2,876 | 539 | 148 | -7% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.