In partnership with
Hello {{first_name | AI enthusiast}},
GPT‑5.6 Sol Ultra posts breakout benchmark scores, Bridgepoint commits €200 million to Skello’s AI HR stack, and Secure.com launches autonomous AI red teaming that hunts exposures continuously instead of quarterly. UN scientists warn on systemic AI risks while fanfiction communities quietly deploy detectors.
Scroll down to catch the signals that matter.
A note from my own desk this week.
I ran the AI governance read I had been building toward, and wrote it up as The AI Governance Convergence, the first of four quarterly reads for 2026 I'm calling Drops.
The short of it: the governance conversation has already moved off the word most 2026 plans are built on. "AI Regulation" is shrinking. "AI Governance" is where the field now tags its work, by about eight to one. And inside that conversation, 12 of the top 20 pains enterprises name are compliance-coded, not model-coded.
Sixty pages, every claim carries its evidence. The single read is €99. The full-year bundle (four reads plus the trackers between them) is €499, the launch price this week.
Read The AI Governance Convergence →
What's in Today?
- 🧠 GPT‑5.6 Sol Ultra Pushes Frontier Reasoning Scores
- 💼 Bridgepoint Leads €200M Bet on Skello’s AI HR Platform
- 🧨 Jailbreak Research Shows AI Safety Defenses Still Lag
- 🛡️ Secure.com Launches Continuous AI Red Teaming for Infrastructure
- 🌍 UN Panel Issues First Global Scientific Assessment of AI Risks
- 📚 Fanfiction Communities Quietly Deploy AI Detection and Face Limits
🧠 GPT‑5.6 Sol Ultra Pushes Frontier Reasoning Scores
OpenAI’s GPT‑5.6 Sol Ultra is emerging as a new reasoning heavyweight, pairing state-of-the-art coding results on Terminal‑Bench 2.1 with stronger scores on biology and cybersecurity benchmarks. Its “Ultra” mode orchestrates specialist sub‑agents, enabling deeper multi‑step problem solving and more robust long‑horizon workflows for advanced users and enterprises.
Expect rapidly escalating competitive pressure on rival foundation models as multi‑agent “Ultra” modes become the new bar for frontier reasoning.
You've seen the AI demos. Viktor does it without you watching.
The AI tool you tried last quarter waited for a prompt, hallucinated a number, then asked if you'd like a summary.
Viktor opened a PR at 2am, rebased it against main, ran your test suite, and posted a note in #eng: "Two flaky tests in payments service, both pre-existing. Recommended merging after fixing them." Then drafted the customer reply for the support ticket the bug created.
That's 619K autonomous actions per day across 20,000+ teams. Not chat replies. Real work shipped to GitHub, Stripe, Linear, Notion, and 3,000+ other tools, from inside Slack and Microsoft Teams.
You don't supervise him any more than you supervise a senior engineer.
SOC 2 certified. Your data never trains models.
"It's what you probably originally thought AI was going to be when you first heard of it in sci-fi movies." Tyler, CEO.
💼 Bridgepoint Leads €200M Bet on Skello’s AI HR Platform
Private equity group Bridgepoint is reportedly set to lead most of a €200 million funding round into French HR software company Skello. The company uses AI to automate workforce scheduling, compliance, and labor forecasting, targeting hospitality, retail, and services where staffing volatility is high and margins are thin.
Expect intensified competition among HR tech players as investors double down on AI‑native workforce optimization platforms.
🧨 Jailbreak Research Shows AI Safety Defenses Still Lag
A detailed essay on the enduring AI jailbreak problem shows how prompt‑engineering attacks and multi‑step exploits still reliably bypass major models’ safety layers. The analysis surveys red‑teaming studies, notes limited effectiveness of current guardrails, and argues that scaling models without robust mitigations amplifies real‑world misuse and compliance risk.
Expect regulators, enterprises, and labs to push beyond superficial guardrails toward formal safety guarantees and continuously updated defense playbooks.
🛡️ Secure.com Launches Continuous AI Red Teaming for Infrastructure
Security firm Secure.com unveiled new autonomous red‑teaming capabilities that use AI agents to probe infrastructure continuously, not just during periodic pen tests. The system chains reconnaissance, exploit simulation, and prioritization, aiming to identify exploitable misconfigurations faster than attackers and shrink the “exposure gap” between deployment and remediation.
Expect CISOs to pivot budgets toward continuous, AI‑driven offensive testing as static security assessments lose credibility.
Two Minutes to Know What Slow Billing Is Costing You
Most SaaS finance teams know their billing process is slow.Most SaaS finance teams know their billing process is slow. Few know what it's costing them.
The Tabs Billing Lag Calculator puts a dollar figure on it in two minutes — benchmarked against top SaaS companies.
🌍 UN Panel Issues First Global Scientific Assessment of AI Risks
The UN’s Independent International Scientific Panel on AI released its preliminary scientific assessment of current capabilities, upside, and systemic risks. The report highlights productivity and health gains but flags concentration of power, labor disruption, biosecurity concerns, and opaque frontier models as strategic governance challenges for governments and multilateral institutions.
Expect this assessment to shape upcoming global AI compacts, export‑control debates, and risk‑based regulatory frameworks.
📚 Fanfiction Communities Quietly Deploy AI Detection and Face Limits
An AI news briefing segment spotlights fanfiction platforms experimenting with AI‑generated content detectors to preserve community norms and originality. Developers describe practical constraints: false positives, adversarial paraphrasing, and limited interpretability when large models flag text without transparent reasoning, raising questions about fairness, moderation legitimacy, and creator trust.
Expect grassroots creator communities to become early testbeds for real‑world AI detection, governance, and interpretability challenges.
What trends are you tracking? Reply with your take or forward it to a colleague shaping AI strategy.