Stop letting anonymous site traffic slip away. Our AI BDR Ava identifies individuals on your website without them providing any contact information and autonomously enrolls them into multi-channel sequences.

She operates within the Artisan platform, which consolidates every tool you need for outbound:

Hire Ava to slash costs & boost productivity.

🚫 Anthropic abruptly cut off Windsurf's access to Claude models with just five days' notice, coinciding with reports of OpenAI's $3 billion acquisition of the AI coding platform. The strategic timing suggests Anthropic is blocking its rival from leveraging Claude's capabilities through the acquisition, even though Windsurf offered to pay for full capacity. While other coding platforms retain Claude access, Windsurf users must now rely on alternatives like Gemini 2.5 Pro or bring their own API keys. This defensive move signals how AI companies are increasingly weaponizing model access as the industry consolidates, creating risks for startups dependent on third-party AI services.

🕵️ OpenAI detected Chinese state-aligned groups exploiting ChatGPT to generate divisive political content and support cyber operations, marking a concerning evolution in AI weaponization. The groups created inflammatory posts about Trump's tariffs, false accusations against activists, and polarized content designed to inflame both sides of US political debates—a sophisticated tactic that exploits societal divisions rather than pushing simple propaganda. While OpenAI has banned these accounts and notes the operations remain limited in reach, the incidents reveal how generative AI tools are becoming strategic assets in state-sponsored influence campaigns. This threat intelligence emerges as OpenAI secures a $40 billion funding round, underscoring the company's growing role as both an AI innovator and a de facto gatekeeper against malicious uses of its technology.

🔒 X updated its developer agreement to ban third-party AI companies from using tweets to train their models, while preserving exclusive access for its own Grok AI system. This move mirrors Reddit's strategy, which recently sued Anthropic for allegedly scraping its site over 100,000 times despite similar restrictions and signed an exclusive training deal with Google. The policy shift highlights a growing trend where social platforms monetize their user-generated content through AI licensing deals rather than allowing free scraping. With Elon Musk's xAI acquiring X for $33 billion in March, the platform now wields its data as a competitive weapon—blocking rivals while using the same content to train Grok, creating a walled garden approach that could reshape how AI companies access training data.

🏗️ Amazon announced a $10 billion investment in rural Richmond County, North Carolina, to build a massive cloud computing and AI infrastructure campus that will create at least 500 technical jobs. The investment targets a region that lost its economic foundation when textile and apparel manufacturing disappeared decades ago, offering a potential model for how AI infrastructure development could revitalize post-industrial communities. Beyond direct employment, the project will modernize critical infrastructure including water systems, wastewater facilities, and fiber optic networks—improvements that benefit the entire community. Amazon's commitment to support local universities and workforce training programs addresses the crucial challenge of transitioning workers from traditional manufacturing to high-tech roles, though questions remain about whether 500 jobs justify the public incentives offered for such a massive investment.

⚠️ The FDA prematurely launched an agency-wide AI system called Elsa that provides inaccurate information about FDA-approved products, despite Commissioner Makary's claims of being "ahead of schedule and under budget." Staff testing revealed the Anthropic Claude-based tool gives completely or partially incorrect summaries, with employees warning that leadership and the Department of Government Efficiency have "overinflated" its capabilities while rushing deployment without proper guardrails. The $28.5 million Deloitte-developed system exemplifies the dangers of prioritizing speed over accuracy in regulatory AI deployment—one staffer bluntly stated that while "Makary and DOGE think AI can replace staff and cut review times, it decidedly cannot." This premature rollout risks compromising the FDA's core mission of ensuring public safety, demonstrating how political pressure for efficiency can undermine the careful validation required when deploying AI in high-stakes regulatory environments.

👨‍💻 Shadow AI has become a critical enterprise risk, with 90% of IT leaders expressing concern about employees using unauthorized AI tools, according to a Komprise survey of 200 IT executives. The research reveals that 80% of organizations have already suffered negative consequences from unregulated AI use, including data leaks and false results, while 13% report financial losses and reputational damage. This shadow AI phenomenon mirrors the earlier shadow IT challenges but with higher stakes—generative AI's ability to process and potentially expose sensitive data creates unprecedented risks. As Krishna Subramanian of Komprise notes, "the cracks are starting to show" as enterprises rush to adopt AI without proper governance, prompting 75% of companies to invest in data management technologies and monitoring tools to regain control over their AI landscape.

📊 Epoch AI operates as a nonprofit research organization dedicated to tracking AI's development trajectory through data-driven analysis and open information sharing, serving as a neutral arbiter in an increasingly polarized field. The three-year-old organization maintains an open-source intelligence program monitoring AI models and hardware, develops standardized benchmarks like FrontierMath for measuring capabilities, and provides independent evaluations through public dashboards—all while deliberately avoiding advocacy positions on whether AI will ultimately benefit society. Their commitment to transparency means most research is shared publicly with policymakers, journalists, and developers, creating crucial infrastructure for evidence-based decision-making about AI. This positions Epoch as an essential counterweight to potentially biased perspectives from organizations with direct financial stakes, though questions remain about maintaining complete neutrality when conducting commissioned research for companies like OpenAI.