At LegalTechTalk London, HaystackID’s John Wilson and Cyril Amarchand Mangaldas’ Komal Gupta examined how deepfakes, business email compromise, and weak cyber governance can converge into multimillion-dollar losses. Their practical message: rehearse response plans, share cyber risk across functions, and use verbal-only challenge phrases before approving high-risk decisions.
At LegalTechTalk, the password is blueberry hot dogs Read More »
GitHub disabled 73 Microsoft repositories after the Miasma campaign re-compromised Azure’s durabletask project, while researchers separately documented Miasma payloads that can detonate when developers open affected repositories in AI coding agents such as Claude Code and Cursor. Here is how the campaign spreads and what security, compliance and eDiscovery teams should do now.
When the worm targets the assistant: Miasma turns AI coding agents into the trigger Read More »
Professional buyers are letting AI build the shortlist before they ever call a vendor. Here is what the research shows about being seen, and how the rules of online visibility are changing.
AI is becoming an advisor, and it is rewriting how buyers find you Read More »
Anthropic expanded Project Glasswing beyond its roughly 50 initial partners to a new cohort of approximately 150 organizations in more than 15 countries, pushing its restricted Claude Mythos Preview model into power utilities, hospitals, telecoms and hardware vendors, and into the records-retention and cyber-insurance questions facing every CISO, general counsel and information governance lead.
At the Dublin Tech Summit, two HaystackID forensics leaders argued that deepfakes have broken the 30-year-old integrity test for digital evidence, and laid out a 90-day plan for proving authenticity under the EU AI Act, NIS2, DORA and the proposed Rule 707.
When you can’t trust the evidence: deepfakes force a forensic reckoning in Dublin Read More »
FutureLaw 2026 closed in Tallinn with the conversation getting harder. Uwais Iqbal’s hard truths from 23,000 adjudications and Chas Rampenthal’s reading of Texas Opinion 705 anchored a day-and-a-half of sessions on predictive justice, productized law, and transnational legal infrastructure.
FutureLaw 2026 closes: hard truths, the billable hour, and what gets built next Read More »
Anthropic’s May 12 Claude for Legal release ships 20-plus MCP connectors and 12 practice-area plugins, re-segmenting the legal AI vendor map and putting setup-interview customization at the center of the buying decision.
Claude for Legal Arrives, and the Legal AI Stack Gets Re-Segmented Overnight Read More »
EU lawmakers struck a provisional political deal on the Digital Omnibus on AI in the early hours of May 7 that, if formally adopted, would defer the AI Act’s flagship high-risk obligations into late 2027 and add a categorical ban on AI-generated CSAM and non-consensual intimate imagery.
EU AI Act Deal Would Delay High-Risk Rules to 2027, Ban Abusive AI Content Read More »
Artificial intelligence is moving from promise to proof in cybersecurity. Organizations using AI extensively in security have cut average breach costs by $1.9 million and shortened breach lifecycles by about 80 days, according to data anchoring a new World Economic
AI in Cybersecurity Moves from Promise to Proof as WEF and KPMG Track Defender Gains Read More »
China’s National Development and Reform Commission has ordered Meta and Manus to unwind a $2 billion AI acquisition closed four months earlier — the first publicly announced foreign-investment prohibition in AI under Beijing’s CFIUS-equivalent security review framework. The decision reshapes diligence and procurement risk for legal-tech buyers, eDiscovery vendors, and any enterprise contracting with AI providers tied to China-origin engineering or training data.
China’s Meta-Manus Block Adds New Risk Layer to Cross-Border AI Diligence Read More »
Stanford’s 2026 AI Index shows AI adoption and investment accelerating even as transparency declines, incidents increase, and governance struggles to keep pace. For cybersecurity, information governance, and eDiscovery professionals, the report signals a more urgent need for defensible oversight, stronger vendor scrutiny, and authenticity-focused workflows.
Stanford’s 2026 AI Index Highlights Rapid Growth and Widening Governance Gaps Read More »
The FTC’s settlement with Match Group over OkCupid’s undisclosed transfer of three million user photos to AI startup Clarifai marks the first federal enforcement action framing AI training data collection as a consumer protection violation — with no fine but a 20-year compliance leash.
FTC’s OkCupid Action Reframes AI Training Data as a Consumer Protection Issue Read More »
