This briefing is informational only and does not constitute legal or financial advice. It reports on public industry developments and does not reference any Tienta client.
This Week in AI
Platform Roundup
- Anthropic (Claude): Claude's text output now carries an invisible watermark by default, a step Anthropic took to comply with the EU's new AI transparency rules. The company says it has no effect on quality, speed, or pricing, and a detection tool is coming soon.
- OpenAI (ChatGPT): OpenAI is expanding its ad pilot in ChatGPT's free tier — ads appear below answers, are labeled as sponsored, and won't show in paid plans. Worth flagging to any staff using the free version for client-facing work: it's now an ad-supported product, not a neutral tool.
- Google (Gemini): Google is rolling out an AI "canvas" directly inside Google Sheets — describe the edit or analysis you want in plain language and Gemini builds it into the spreadsheet. Useful if your team already lives in Sheets for budgets or trackers.
- Microsoft (Copilot): Microsoft has started merging its consumer and business Copilot apps into a single "Copilot" app. In the process, it's retiring Copilot Podcasts, Group Chat, and Deep Research (consumer version) on August 18 — if your team uses any of those, save what you need before then.
- xAI (Grok): xAI is facing a newly filed lawsuit alleging its image-editing tool was used to generate child sexual abuse material, adding to existing litigation over the platform's content safeguards — a reminder that vendor legal exposure is worth tracking alongside features when evaluating any AI platform.
Beyond the Platforms
Stripe buys AI "model router" OpenRouter for over $7 billion
Payments company Stripe finalized a deal this week to acquire OpenRouter, a startup that lets developers switch between more than 400 AI models through one connection point, at more than five times the valuation OpenRouter held just three months ago.
Why it matters: as the company that already processes a lot of the internet's payments moves into AI infrastructure, expect AI model access and billing to become more bundled with the payment tools many small businesses already use.
Anthropic raises its own risk rating for Claude
In its twice-yearly safety report, Anthropic raised its assessment of the risk that its models could take seriously harmful, misaligned actions from "very low" to "low," citing broader uncertainty following recent security-testing disclosures rather than any single new failure.
Why it matters: this is a case of a vendor volunteering a more cautious rating about its own product rather than being caught out — a reasonable factor to weigh when choosing which AI providers to trust with sensitive work.
Industry Spotlight
For Legal Practices
NYC Bar committee calls for a national AI rulebook
A New York City Bar Association committee published a policy paper this week concluding that AI tools "may assist legal work but cannot substitute for professional legal judgment," and called for AI-specific standards to be built directly into the national Model Rules of Professional Conduct rather than left to a patchwork of state opinions.
Why it matters: a unified national standard would simplify compliance for firms operating across state lines — worth watching as a possible reset of today's fragmented guidance.
AI is driving more self-represented litigants into court — without more wins
New reporting this week finds that AI tools are fueling a rise in pro se (self-represented) court filings, but that increase hasn't translated into better outcomes for those filers.
Why it matters: courts and opposing counsel are seeing more AI-assisted filings from unrepresented parties, which can add friction and unpredictability to litigation timelines regardless of case merit.
Clients want AI-driven pricing changes; most firms haven't made them
Thomson Reuters' latest Future of Professionals research finds 71% of in-house legal professionals expect their outside firms to adjust pricing as AI usage grows, but only 28% of firms report having actually changed their pricing structure.
Why it matters: this gap is becoming a competitive opening — firms that move first on AI-adjusted pricing may be better positioned as client expectations keep shifting.
For Financial & CFO Firms
Accounting teams' AI ambitions are outrunning their execution
A new study finds 85% of accounting teams have made AI a strategic priority, but only 10% report using it extensively — a gap the report ties to manual work still bogging down teams despite stated priorities.
Why it matters: if your firm has declared AI a priority on paper, this is a good week to check whether that's translated into actual day-to-day use, or just a strategy slide.
Accounting firms are starting to rethink pricing around AI
BILL's latest Accounting Firm AI Ambition survey finds firms are restructuring client advisory service offerings and pricing models to capture the economics AI has enabled, rather than simply doing the same billable work faster.
Why it matters: as AI compresses the time routine compliance work takes, firms that keep billing by the old hour risk giving away the efficiency gains for free — worth a look at how your own fee structure holds up.
Banking industry pushes back on state-by-state AI rules
The American Bankers Association responded this week to a congressional request for input on AI policy, detailing the sector's existing risk-management and cybersecurity practices and calling for a single federal AI framework that would preempt the growing patchwork of state rules.
Why it matters: if your firm operates in multiple states, the outcome of this federal-versus-state debate will shape how many different AI compliance regimes you eventually have to track.
On the Horizon
Finance chiefs are moving fast on AI but not on the guardrails around it. Deloitte's latest CFO Signals survey found that while 93% of surveyed finance leaders now use AI across multiple functions, only 43% feel confident in their organization's AI governance — the rest are only "somewhat" confident or less.
Why get ahead of this now: governance is far easier to build before an AI-related mistake forces the issue than after one. A short internal review of who's using which tools, on what data, is cheap insurance against a costly retrofit later.
Client expectations on AI-adjusted pricing are running well ahead of what most professional-services firms have delivered (see the Thomson Reuters figures above, which apply just as much to accounting as to law). Multiple analysts now frame this gap — high ambition, low execution — as the defining tension in professional services for the next year.
Why get ahead of this now: firms that can clearly show clients where AI has actually changed their costs and pricing will have a real story to tell; firms that can't may face harder fee conversations as the gap becomes more visible to clients.
Try This This Week
If your team already uses Google Sheets for budgets, trackers, or client data, try Gemini's new in-Sheets canvas feature this week: describe what you want built or analyzed in plain language instead of writing formulas by hand. Treat the output the same way you'd treat any AI-generated number — check a handful of the calculated cells against your own math before you rely on it. It's a fast way to see where AI can save real time in a tool you're already using, without adopting a whole new platform.