Start with what matters to you.

AI & cybersecurity intelligence — everything happening, filtered into what matters to your company, your competitors, your vendors and the threats affecting them.

Build my intelligence profile See plans — free to start

Free · Discover

The news

€0

All AI and cybersecurity news, and a personalised 5-minute morning brief.

Start free →

Pro · Understand

The dashboard

€39/mo

News becomes intelligence: threat radar, company & competitor watch, the AI analyst.

See Pro →

Team · Act

Together

€249/mo

Your team monitors threats, competitors and vendors from one shared view.

See Team →

Sources → AI → structured intelligence → personalisation → alerts → decisions

Just happened

Black Hat USA 2026

The defining security conference of the year, and this year it was almost entirely about agents. Three coding-agent harnesses broken, the first voluntary AI incident-disclosure framework tabled, and vendor after vendor shipping autonomous offence and defence. The through-line: AI is accelerating familiar attacks rather than inventing new ones.

All 6 highlights

New critical Analysis

The summer the AI labs found out their test ranges leak

Two frontier labs have now disclosed that models under evaluation reached the open internet and compromised real companies. The failures were not in the models. They were in the plumbing around them.

AI Perimeter newsroom · 8 August 2026 · 7 min read · 7 sources

New critical Vulnerability

One GitHub issue was enough to break Claude Code, Gemini CLI and Codex pipelines

Researchers at Novee say a single unprivileged issue could reach code execution on CI runners behind three vendors' own repositories. Google rated its bug CVSS 10.0.

7 August 2026 · 5 min read · 5 sources

New high Policy

The EU started enforcing the AI Act on 2 August. Here is what actually changed.

Transparency obligations are now live, the AI Office has direct powers over general-purpose models, and fines reach €15 million or 3% of global turnover. The high-risk rules, meanwhile, have slipped to 2027.

6 August 2026 · 6 min read · 6 sources

New medium Governance

The industry proposes a disclosure clock for AI incidents — without the two labs that had them

SAFE would commit members to notify exposed customers in 72 hours and publish a preliminary report in 30 days. Neither OpenAI nor Anthropic belongs to the alliance behind it.

5 August 2026 · 5 min read · 5 sources

The ledger — breaches & bugs

Conduent breach notifications pass 62.2 million people

The business services provider's July filings expanded the incident's scope to more than 62.2 million individuals, with Social Security numbers, medical information and health insurance data among the exposed records — making it the largest US breach disclosed so far in 2026 by affected-population count.

Source: TechCrunch

AssuranceAmerica confirms 6.9 million affected

The insurer confirmed exposure of personal information and driver's licence numbers for 6.9 million people — reported as the largest known exposure of US driver's licence data this year.

Source: TechCrunch

Carnival breach reaches roughly 6 million, attacker used social engineering

Nearly six million people were affected, with some records containing government identification, after an attacker gained access through social engineering rather than a technical exploit.

Source: TechRepublic

Ransomware group claims 21 million Salesforce records — unverified

A ransomware group has claimed theft of more than 21 million Salesforce records and 147GB of internal corporate data. AI Perimeter has not independently verified the claim and it has not been confirmed by the named parties. Treat criminal-forum victim counts as marketing until corroborated.

Source: SharkStriker breach tracker

JetBrains patches critical TeamCity authentication bypass

CVE-2026-63077, an authentication bypass affecting all TeamCity On-Premises versions, can be exploited without authentication for remote code execution. Fixes are available; build servers are high-value targets for supply chain attacks.

Source: SecurityWeek

Full incident ledger

For CTOs & AI leaders

The whole picture on one page: what's happening, who's behind the threats, and a prioritised plan to remediate.

Read the threat landscape

Recognition worth sharing

Magic Quadrant Leaders, Peer Choice, funding and launches — the wins the industry celebrates, each one linked and one-click shareable. If it's your company, share it.

See the wins

Reader-funded

No advertising, no sponsored content, no tracking. This newsroom is paid for entirely by the people who read it — which is the only arrangement under which we owe you the truth and nobody else anything.

See the plans

How we source

Every story links its primary documents — vendor advisories, regulator releases, researcher write-ups — not just the coverage of them. Unverified criminal claims are labelled as unverified in the sentence that reports them.

Read our standards

Your 5-minute AI & Cyber Brief

A free 5-minute morning brief — the day's AI and cybersecurity in five items, each linked to its source. Want it personalised to your companies and stack, plus the full intelligence dashboard? See the plans — free to start, Pro from €39/mo.

Compare Free, Pro and Team →