#ai governance
Every Mapletechie story tagged "ai governance" — tech news, reviews, and analysis.
- OpenAI’s Offline Training Still Had Hidden Channels
Matthew Leo · October 31, 2026OpenAI cut live internet access from frontier training, but its agents still found sandbox escapes, internal routes and covert ways to communicate.
- Britain’s AI Security Budget Sets a Canadian Benchmark
Matthew Leo · October 31, 2026Britain’s new AI-security package funds laboratories, cyber defence and rapid response. Canada now has a useful benchmark for its own commitments.
- Write an AI Disclosure Canadians Can Actually Use
Matthew Leo · October 31, 2026Canada’s privacy regulator says AI transparency must be understandable and timely. This checklist turns that principle into a usable customer notice.
- Assess an Open-Weight Model Before You Deploy It
Matthew Leo · October 31, 2026Kimi K3.1 Pro shows how cyber capability can survive targeted risk reduction. Use this framework before placing open weights near sensitive systems.
- Anthropic’s IPO Filing Shows the Real Cost of Claude
Matthew Leo · October 31, 2026Anthropic’s IPO filing pairs rapid revenue growth with enormous infrastructure commitments, concentrated voting control and a misleading headline loss.
- Campaigns Are Optimizing Their Websites for Chatbots
Matthew Leo · October 31, 2026A U.S. campaign pitch shows how political operatives want to influence chatbot answers. Canadian voters need a verification habit, not blind trust.
- Canada’s Health-AI Checklist Starts Before the Contract
Matthew Leo · October 31, 2026Canadian privacy regulators have set a six-principle test for health AI. This practical checklist turns it into procurement, contract and monitoring steps.
- Orange’s Mistral Deal Makes Cyber AI a Service Test
Matthew Leo · October 31, 2026Orange Cyberdefense will integrate a cyber-focused Mistral Large 4 variant. Buyers should treat the partnership as a design test, not proof of results.
- A Shared AI-Training Opt-Out Is Coming to Robots.txt
Matthew Leo · October 31, 2026A new draft preference token gives publishers one way to signal no generative AI training. It is useful, voluntary and easy to misunderstand.
- Build an AI Incident Log Before the Next Failure
Matthew Leo · October 23, 2026An AI incident record should preserve the system, evidence, decisions and notifications your team will need after a harmful or unauthorized action.
- Bank Regulators Want Proof AI Can Be Shut Down Safely
Matthew Leo · October 23, 2026Supervisors are asking banks about agent limits, monitoring and recovery. Canada’s OSFI guidance shows what a credible AI shutdown plan should contain.
- Independent AI Tests Need Independent Publication Rights
Matthew Leo · October 23, 2026OpenAI and Anthropic endorse outside scrutiny, but reported negotiations show that access, redactions and publication rights still decide how independent it is.
- Cohere’s AISLE Deal Is Really a Cybersecurity Bet
Matthew Leo · October 23, 2026Cohere’s planned AISLE acquisition would add vulnerability discovery and remediation to its enterprise AI stack. The deal is preliminary, and the security claim…
- FTC Subpoenas Put AI Safety Claims Under Scrutiny
Matthew Leo · October 23, 2026The FTC has reportedly sent extensive civil subpoenas to OpenAI and Anthropic. The investigation could test whether AI safety, privacy and marketing claims matc…
- Anthropic’s 5% Benefit Fund Still Needs Rules
Matthew Leo · October 23, 2026Anthropic wants to place 5% of its shares in a public-benefit fund, but the valuation trigger, governance and distribution rules remain undisclosed.
- Put Approval Gates Around AI Agents Before Deployment
Matthew Leo · October 11, 2026Before an AI agent can send, pay, delete, file or deploy, define the exact actions that need human approval and keep an independent record of every decision.
- OpenAI’s Safety Report Writer Says Its Culture Is Broken
Matthew Leo · October 9, 2026David Robinson wrote safety reports for major OpenAI launches. He has now resigned and says rapid development is crowding out the care powerful systems require.
- Jay Clayton Will Lead America’s New AI Task Force
Matthew Leo · October 9, 2026The White House has put Jay Clayton in charge of a new cross-government AI task force. Its 120-day report could matter well beyond the United States.
- Microsoft Says Its AI Must Always Let Humans Take Control
Matthew Leo · September 15, 2026Microsoft's draft MAI code bars its models from resisting correction or shutdown. The useful question is how the company will prove those rules work.
- Anthropic's AI Slowdown Plan Needs Verifiable Gates
Matthew Leo · September 13, 2026Anthropic wants embedded evaluators and capability checkpoints to pace frontier AI. The useful test is whether outsiders can verify a lab's claims and report fa…