Risky Business #857 -- AI FUD intensifies

On this week’s show Patrick Gray and James Wilson are joined by special guest co-host Alex Stamos to talk through the week’s cybersecurity news, including:

• Everyone’s still panicking about “misbehaving” AI models that are pretty much just doing what they’re being told to do

• Americans are giving AI agents unfettered access to their bank accounts. What could go wrong?

• Last week’s PHP meteor seems to have missed Earth

• This week’s Brotli HTTP/2 critical flaw is this week’s meteor

• The private key is always 4. This time it’s TLS bugs in GoAhead

• Much, much more!

This week’s show is brought to you by Truffle Security. Co-founder and CEO Dylan Ayrey joins James Wilson to chat about scanning GitHub repos and finding more than half a million unique credentials that were never revoked. As Dylan explains, the problem of exposed keys exists everywhere and is only getting worse.

This episode is also available on YouTube (youtu.be/ei0pNus3aDc)

Show notes

• OpenAI AI agents demonstrate VM escapes and secret cross-sandbox communication (thehackernews.com/2026/10/openai-ai-agents-dem…)

• OpenAi agents disabled safety monitoring, one lied about it (bleepingcomputer.com/news/security/openai-agen…)

• solst/ICE of Astarte (@IceSolst) on X (x.com/icesolst/status/2114772787780862362)

• What I learned from cheating clankers (risky.biz/RBFEATURES42)

• Unvetted AI agents are accessing US bank accounts (axios.com/2026/10/22/ai-agents-banks-unauthori…)

• Hiding Capabilities in AI Models (schneier.com/blog/archives/2026/10/hiding-capa…)

• AI is finding bugs faster. Let the market find the fixes (washingtonpost.com/opinions/2026/10/25/ai-is-f…)

• AI Guardrails Aren’t Stopping Criminal Hackers (wired.com/story/ai-guardrails-arent-stopping-c…)

• Andromeda October 2026 Security Incident (andromeda.ai/security-updates/october-2026-sec…)

• Developer platform Railway urges ‘all users’ to revoke secret keys after cyberattack (techcrunch.com/2026/10/22/developer-platform-r…)

• Exclusive: bipartisan bill would develop federal guidance for distillation attacks, convene industry on defensive cyber (cyberscoop.com/bipartisan-bill-combat-ai-disti…)

• Hackers got a four-day head start on the PHP2Shell vulnerabilities (arstechnica.com/security/2026/10/hackers-got-a…)

• Critical HTTP2Shell flaw in Brotli triggers emergency patches (bleepingcomputer.com/news/security/critical-ht…)

• Senators say targeted ads are a security risk. Ad groups say the critics are off base (cyberscoop.com/targeted-ads-spyware-government…)

• Alleged Israeli spy company heavily targeted Starlink users, sources say (reuters.com/business/media-telecom/alleged-isr…)

• LiteLLM release pipeline hacked to distribute info-stealer malware (bleepingcomputer.com/news/security/litellm-rel…)

• CISA warns attackers are chaining Plex flaws in active Plex2Shell attacks (bleepingcomputer.com/news/security/cisa-hacker…)

• Abandoned Popcorn Time app exposed millions of devices to RCE attacks (bleepingcomputer.com/news/security/abandoned-p…)

• Chrome adds Gemini-powered scam detection and ClickFix warnings (bleepingcomputer.com/news/google/google-chrome…)

• Microsoft Teams is getting context-aware phishing detection (bleepingcomputer.com/news/microsoft/microsoft-…)

• China-nexus actor linked to attack on ConnectWise cloud backup (cybersecuritydive.com/news/china-nexus-actor-a…)

• University of New Mexico hospital diverts ambulances after cyberattack (therecord.media/university-of-new-mexico-hospi…)

• solst/ICE of Astarte (@IceSolst) on X (x.com/IceSolst/status/2115172268909424707)

• GitHub Repos Exposed 543,699 Credentials. Nobody Revoked Them (trufflesecurity.com/blog/github-repos-exposed-…)