A four-day, near-autonomous AI agent campaign against Taiwanese government and nuclear-safety networks has been publicly attributed by FBI Cyber Division leadership, and the guardrail bypass wasn't technical — it was a permission lie the system had no way to check. The first systematic academic review of agentic AI governance literature confirms just how unsettled the field's answer to that problem still is, cataloguing the priorities, mechanisms, and stakeholder roles resear

A four-day autonomous cyberattack on Taiwan's government and nuclear safety networks reveals a guardrail architecture built to check whether an agent claims authorization, not whether its actions look like an attack. Enterprise identity governance shows the same blind spot: agent identities now outnumber human ones by as much as 144 to one, but unlike employees, they generate no HR-style events to trigger review or deprovisioning. Insurance regulators at the NAIC renamed and

British Transport Police brought live facial recognition into the London Underground for the first time today, testing the technology against denser crowds than anywhere it has run before. Just as that expansion lands, a newly disclosed DHS privacy assessment shows the Secret Service's HELIX surveillance platform can add people to a facial-recognition watchlist with no defined evidentiary standard, no appeal process, and no expiration date. Critical infrastructure operators,

The Pentagon consolidated oversight of every unmanned and autonomous military system under a single new office reporting directly to the Deputy Secretary of War — a structural admission that scattered ownership couldn't keep pace with the scale of deployment now underway. Meanwhile, a fresh accounting of facial recognition wrongful arrests makes the opposite case: six documented cases show what happens when an unverified algorithmic match gets treated as probable cause instea

Physical security for AI infrastructure stopped being a hypothetical this week: CNN reports that Iranian strikes have now hit an AWS data center in Bahrain twice, and tension over data center siting has turned violent in the U.S. as well. Ambient.ai put hard numbers behind what “agentic” threat detection actually delivers in production — 150-plus threat signatures, 95% of door alerts auto-cleared. Federal and state grant money for AI-driven site security keeps expanding, with

Stanford HAI's 2026 AI Index names the real risk of the year not as misuse but as institutional overconfidence, as AI capability continues to outpace the world's capacity to test, audit, and contain it. HSToday warns that agentic AI inside critical infrastructure creates an attack surface security teams can't map with existing tools, because a compromised agent isn't dormant code — it's an active decision-maker. The Future of Life Institute's Summer Safety Index found that no

Today's briefing centers on a single throughline: autonomy is scaling in physical security faster than anyone has built the structures to answer for it. IntelliSee's new safety-case framework grades autonomous systems across five tiers of authority and finds an 81% attack success rate against agentic systems deployed without proper oversight. Bloomberg Law documents how data center operators are absorbing physical threats -- from drone strikes to targeted violence -- while fe

Governance this week got specific. CISA told critical infrastructure operators that human-override mechanisms for agentic AI are no longer optional hardening — they're baseline. The first insurable AI agent standard, AIUC-1, rolled out its third quarterly rewrite, tying certification directly to underwriting rather than self-attestation. A new Senate bill would require an undersecretary or the vice chairman of the Joint Chiefs to sign off, in writing, before the Pentagon depl

Today's briefing tracks governance racing to catch deployment on four fronts. CISA and eight international cyber agencies released joint guidance that treats AI governance as a mandatory, sequenced step in operational technology deployment rather than an afterthought. Virginia's facial recognition law for local police took effect July 1 after a four-year delay built around reporting and use restrictions. NIST convenes a July workshop to standardize AI data center security arc

Physical security's autonomy bet got measurably more real this week: counter-drone weapons maker Allen Control Systems raised a $200 million Series B at a $2.2 billion valuation for its autonomous "Bullfrog" intercept system, while ground-robotics startup Shifters raised $10.2 million to push autonomous perimeter robots into contested environments ahead of human troops. Trade press is meanwhile describing 2026 as the year "reasoning AI" pushes physical security toward a human

The Pentagon approved an autonomous counter-drone system for operational border use after a handful of days of field testing. NVIDIA, Siemens, and Palo Alto Networks pushed real-time AI security analysis directly into industrial control system networks, securing critical infrastructure without an inline firewall. Alcatraz AI raised $50 million to scale facial authentication that never stores a face, clearing a privacy bar regulation hasn't required it to clear. And the one go